rat embryonic ectoderm as renal isograft

J. Embryol. exp. Morph. 94, 1-27 (1986)

Printed in Great Britain © The Company of Biologists Limited 1986

REVIEW ARTICLE

Rat embryonic ectoderm as renal isograft

ANTON SVAJGERInstitute of Histology and Embryology, Faculty of Medicine, University of Zagreb,Salata3, P.O. Box 166, 41001 Zagreb, Yugoslavia

BOZICA LEVAK-SVAJGER AND NIKOLA SKREBInstitute of Biology, Faculty of Medicine, University of Zagreb, Salata3, P.O. Box166, 41001 Zagreb, Yugoslavia

INTRODUCTION

Experimental results obtained many years ago revealed that during gastrulation(with the primitive streak and the mesoderm formation as distinct features) theearly rodent embryo undergoes essential changes in its response to extrinsicteratogens (Russell & Russell, 1954; Wilson, 1954; Skreb, 1961; Skreb & Bijelic,1962; Skreb & Frank, 1963). It has also been shown that the ultrastructural,histochemical and biosynthetic features of the embryo are subject to substantialchanges during this period (Solter, Damjanov & Skreb, 1970, 1973; Dziadek &Adamson, 1978; Bode & Dziadek, 1979; Wartiovaara, Leivo & Vaheri, 1979;Jackson et al. 1981; Franke et al. 1982a, b). This suggests a restriction of devel-opmental capacities (i.e. the loss of the capacity of regulation) in groups ofembryonic cells at this developmental stage.

According to the current concept, the initial cell population from which thisrestriction starts, resides within the embryonic ectoderm of the pregastrulation orpreprimitive streak embryo (primitive or primary ectoderm). The first exper-imental evidence in favour of this view was contributed by Grobstein (1952).A more rigorous analysis has been made possible by the improvement of thetechnique of separation of germ layers (Levak-Svajger, Svajger & Skreb, 1969;Svajger & Levak-Svajger, 1975). The developmental potential of isolated germlayers of the rat embryo has been tested by grafting them under the kidney capsuleof adult syngeneic animals and by histological analysis of the resulting teratomas(Skreb & Svajger, 1975; Skreb, Svajger & Levak-Svajger, 1976; Svajger, Levak-Svajger, Kostovic-Knezevic & Bradamante, 1981).

The space between the kidney capsule and the parenchyma of the adult kidneyoffers a suitable environment for growth and differentiation of both the whole rategg-cylinder and the separate germ layers (Skreb & Svajger, 1975). However, evenin this favourable ectopic site the transferred ectoderm undergoes considerable

Key words: rat embryo, renal isograft, ectopic grafts, kidney capsule.

2 A. SVAJGER, B. LEVAK-SVAJGER AND N. SKREB

changes in its morphogenetic behaviour when compared with its normal de-velopment in situ (Svajger et al 1981). Here we review the essential results fromexperiments with renal isografts of rat embryonic ectoderm, compare them withthe relevant results obtained in other experimental systems and give a criticalevaluation of conclusions in the light of the atypical features which may occur inectopic grafts.

TERMINOLOGY

Several synonyms have been used to designate the three germ layers: ectoderm(epiblast, ectoblast, ectophyll), mesoderm (mesoblast) and endoderm (hypoblast,endoblast, endophyll). The terms epiblast and hypoblast are commonly used forthe primitive germ layers of the early chick embryo, while the terms ectoderm andendoderm are more common in mammalian embryology. The parallel or evencombined usage of these sets of terms is confusing. It is therefore high time todecide to use a strictly unique terminology, at least for the mammalian embryo. Itis important to distinguish between germ layers before gastrulation {the primitiveor primary ectoderm and endoderm) and after the completion of gastrulation{definitive ectoderm, mesoderm and endoderm). The criteria are given in Table 1.Without the adjectives 'primitive' or 'primary' the terms ectoderm, mesoderm andendoderm (endo = ento) mean definitive germ layers. In connection with themesoderm (which lacks a topographically analogous cell layer before gastrulation)the adjectives 'primary' and 'secondary' are sometimes used to designate the

Table 1. Survey of rat embryonic germ layers

Germ layer

Primitive (primary)ectoderm

Primitive (primary)endoderm

(Definitive)ectoderm

(Definitive)endoderm

Mesoderm

Precursor cellpopulation

Inner cell mass

Primitive ectoderm

Stage ofappearance

Blastocyst stage

Gastrulation

Developmentalsignificance

Undergoes gastrulation andgives rise to definitiveembryonic germ layers

Contributes to the extra-embryonic endoderm ofthe visceral and parietalyolk sac

Undergoes neurulation andgives rise to neural tissue,epidermis, other 'ecto-dermal epithelia' andneural crest

Differentiates into thesurface and glandular epi-thelium of the primitive gut

Undergoes partial segmen-tation and gives rise tomesenchymal and epithelial'mesodermal derivatives'

Rat embryonic ectoderm as renal isograft 3

developmental stages of the mesenchyme as the first and most widespread tissuetype of mesodermal origin. The mesodermal cells leave the primitive streak in theform of the primary mesenchyme, which then consolidates into provisory epi-thelial structures (somites, intermediary mesoderm, lateral plates). These in turnpartly dissociate to form the secondary mesenchyme (sclerotome, dermatome,myotome, renal and lateral plate mesenchyme) (Hay, 1968).

The terms proximal and distal (ectoderm) refer to the position in relation to theamnion and other extraembryonic membranes, and the terms anterior {frontal),posterior and lateral refer to the position in relation to the embryonic axis (seePoelmann, 1980ft; Snow, 1981; Beddington, 1983a; Tarn & Meier, 1982; Tarn,1984).

The embryonic germ layers are continuous with those which make the extra-embryonic membranes. The latter will not be considered in this text.

In order to simplify the explanation of some results reviewed here, we willsometimes use the adjectives 'ectodermal', 'endodermal' and 'mesodermal' withquotation marks to designate the tissues or cells which are supposed to derive fromthe corresponding definitive germ layers in situ. However, in view of the atypicalmorphogenetic events which occur in renal isografts of the isolated primitiveectoderm, it is actually meaningless to speak of the germ layer origin of tissueconstituents of the resulting teratomas. In general, the terminological designationsof the germ layers will be used here in their classical, topographical meanings, which donot imply their actual potencies for tissue differentiation in situ and in variousexperimental conditions.

MANIPULATION OF EMBRYOS

Isolation of germ layers includes the treatment of embryonic shields withproteolytic enzymes (0-5 % trypsin + 2-5 % pancreatin in calcium- and mag-nesium-free Tyrode's saline at +4°C for 15-30 min) and microsurgery withelectrolytically sharpened and polished tungsten needles (Levak-Svajger et al.1969; Svajger & Levak-Svajger, 1975; see Snow, 1978 for modifications). Theisolated pieces are transferred with a specially designed glass pipette under thekidney capsule of syngeneic adult male rats. Teratomas are fixed after 15-30 daysand processed for routine histology.

Isolation of germ layers is the crucial step in the procedure because anincomplete separation would render any reasonable interpretation of the resultsimpossible. A clean separation of 'uncontaminated' germ layers is facilitated bythe existence of enzyme-susceptible basement membranes between them (Pierce,1966; Adamson & Ayers, 1979) and by the inherent tendency of the ectoderm torearrange its cytoskeleton and to invaginate during neurulation. The initialdetachment of the ectoderm usually occurs spontaneously after treatment withenzymes. Any subsequent surgery can be controlled by careful inspection duringmanipulation under the operating microscope. The separation of germ layers isonly impossible in the region of the primitive streak, where the ectoderm lacks


a basement membrane (see Mitrani, 1982 and Sanders, 1984 for avian embryos)and continuity exists between mesodermal, endodermal and ectodermal cells(Poelmann, 1980a).

EXOGENEOUS INFLUENCES UPON THE GRAFT

In this experimental system the embryonic ectoderm is subject to atypicalexogeneous influences during isolation, grafting procedures and development inthe host. These are: (a) interruption of the supply of oxygen and nutrients byuterine blood circulation, (b) influence of new environmental constituents (saline,serum, enzymes), (c) surgical trauma, (d) loss of connection with extraembryonicmembranes and of contact with extraembryonic fluids, (e) altered physical(spatial) conditions after transplantation, (f) reduced supply of oxygen andnutrients before full vascularization of the graft, and (g) putative interactiveinfluences from the host tissues. The effects of these factors on the growth,morphogenesis and differentiation of the isolated ectoderm as a renal isograft arepoorly understood or completely unknown.

Isolation and transfer of the ectoderm

Proteolytic enzymes act by digesting the glycoproteins (laminin and fibronectin)of the basement membrane (Zimmermann, Merker & Barrach, 1982). Althoughthe exposure to enzymes is not harmless to the components and the properties ofthe cell surface (Waymouth, 1974; Maslow, 1976), the cells are subsequently ableto restore their basement membrane and other pericellular materials (Osman &Rush, 1981) and to retain a high percentage viability (Dziadek, 1981). In the lightof evidence for rapid wound healing in rat embryos (Smedley & Stanisstreet, 1984)and of an efficient regulation of extensive random cell loss in the mouse embryo(Snow & Tarn, 1979), the surgical trauma during separation and transfer is notlikely to influence the developmental capabilities of the ectoderm.

Influence of the new environment

Prior to isolation the free and the basal surfaces of the embryonic ectoderm arein contact with the amniotic fluid and the basal lamina respectively. After transferto an ectopic site the early embryonic cells are surrounded by mature tissues of thehost. An influence of this new, atypical environment on the grafted ectodermcannot be absolutely excluded. It is, however, highly improbable that thisinfluence is of a classical inductive type. It has been demonstrated that the basicstructure of the renal capsule is essentially the same in various mammalian species.It consists of an outer connective tissue layer, a layer of atypical smooth musclecells with adrenergic innervation and vascularized loose connective tissue adjacentto the renal parenchyma (Bulger, 1973; Kobayashi, 1978). These structures areunlikely to act as embryonic inducers of differentiation. Differently structuredhost tissues: the iris of the rabbit (Levak-Svajger & Skreb, 1965), the capsule ofthe testis (Diwan & Stevens, 1976) and the cheek pouch of the golden hamster


(Damjanov, 1978) also allow a wide range of various tissue differentiation ingrafted egg-cylinders. On the other hand, even in a 'controlled environment'in vitro non-specific constituents of the substratum and the culture medium canelicit the expression of atypical phenotypes in cultured tissue (differentiation ofcontracting striated muscle fibers from pituitary cultures, Brunner & Tschank,1982).

As pointed out previously (Levak-Svajger & Svajger, 1974), the great diversityof differentiation and organotypic tissue associations in teratomas is unlikely to beinduced by the host tissues. The presence of tissue derivatives of particular germlayers varies regularly in relation to the original germ-layer composition of thegraft. It is therefore most probable that the final composition of the grafts reflects agood deal their initial developmental capacities.

The influence of the host tissues might, however, be of a non-inductive nature.This concerns mainly the proper and quick incorporation and vascularization ofgrafts which are regularly achieved in a renal site but much less so in an intraocularone (Skreb & Levak, 1960). The influence of the host's sex hormones (malerecipients!) has not been observed since the appropriate target organs (exceptprostatic rudiments and occasional praeputial glands) have never developed ingrafts.

Also in explants of rat embryonic shields cultivated in vitro the quality of tissuedifferentiation has been shown to vary with respect to the composition and thephysical features of the in vitro environment as well as to the frequency of renewalof the culture medium (Skreb & Crnek, 1977,1980).

MORPHOGENETIC BEHAVIOUR OF THE EMBRYONIC ECTODERM AS A RENAL

ISOGRAFT

It has been demonstrated that during normal gastrulation in situ of both chickand mouse embryos the primitive ectoderm is subject to temporally and spatiallycoordinated movements of individual cells in migrating cell streams which use theprimitive streak as the passageway to their final destinations (Solursh & Revel,1978). Approaching the primitive streak the cells of the primitive ectoderm losetheir basement membrane (Mitrani, 1982; Sanders, 1984, chick embryo), acquire abottle-like shape (Tarn & Meier, 1982, mouse embryo) and move downwards toincorporate into the definitive endoderm and mesoderm. At the edges of theprimitive streak the cells display an active blebbing of the basal surfaces (Sanders,1984). This is a common feature in embryonic epithelial cells which have lost theirbasement membranes (Lieb & De Paola, 1981; Osman & Ruch, 1981; Sugrue &Hay, 1981).

In our experiments isolated rat embryonic ectoderm has been transferred underthe kidney capsule without the basement membrane which in situ maintained itsepithelial structure and regulated the migration of the mesenchyme through theprimitive streak (Hay, 1973; Sanders, 1984). This was very probably the majorcause of atypical and irregular morphogenetic behaviour of the ectoderm in


isografts, which was described in detail by Svajger et al. (1981). The essentialcourse of events comprises the breaking up of the coherent epithelial structure ofthe ectoderm and its conversion into an unorganized, mesenchyme-like cell mass.As a secondary event, rosette-like, cystic or tubular structures form out of thismesenchyme. They probably represent the rudiments of the future epithelialstructures within the teratomas. Another form of mesenchyme production hasbeen observed in grafts of head-fold-stage embryonic ectoderm: protrusion of cellgroups beyond the basal boundary of the epithelium, which is strongly reminiscentof the initial migration of neural crest cells in situ (Vermeij-Keers & Poelmann,1980). Even during normal development of the mouse embryo in situ, at the stagesprior to neurulation, small focal disruptions of the epithelial organization wereobserved in the lateral ectoderm of the mouse embryo (i.e. beyond the primitivestreak region, Poelmann, 1980a).

Dissociation of the epithelium into amorphous cell aggregates was also observedin renal isografts of the caudal end of the mouse embryo (Tarn, 1984), in testiculargrafts of 2-cell mouse eggs (Stevens, 1968) and in cultures of chick digestive tractepithelium (Sumiya, 1976). Although considered here as atypical in comparisonwith normal gastrulation in situ, this basic morphogenetic mechanism is at work invarious developmental processes in vertebrates: the formation of the neural crest(Vermeij-Keers & Poelmann, 1980) and of the secondary mesenchyme (Hay,1968), the formation of the tail bud (Bijtel, 1936; Jolly & Ferester-Tadie, 1936;Peter, 1941; Meier & Jacobson, 1982; Nakao & Ischizawa, 1984; Schoenwolf,1984), and partially in the formation of some cranial peripheral ganglia (AriensKappers, 1941; Verwoerd & van Oostrom, 1979; D'Amico-Martel & Noden,1983). Even regressive processes in embryogenesis, such as the disappearance ofthe Miillerian ducts in the male mouse embryos, may involve this mechanism(Trelstad, Hayashi, Hayashi & Donahoe, 1982; Paranko, Pelliniemi & Foidart,1984). On the other hand, the formation of mesenchyme through an analogoue ofthe primitive streak is always attained when the early mouse embryo (or theembryonic body) is allowed to develop in vitro without disrupting its extra-embryonic cavities and membranes (Wiley & Pedersen, 1977; Wiley, Spindle &Pedersen, 1978; Libbus & Hsu, 1980; Uno, 1982).

The transformation of epithelial cells into mesenchymal ones is a phenomenonof general importance in developmental biology. It can be induced in someembryonic epithelia by appropriate culture conditions (Hay, 1978; Greenberg &Hay, 1982). It is not yet clear whether all epithelia retain an inherent potency tobecome mesenchymal cells under certain circumstances, nor what constitute thecontrol mechanisms of epithelial and mesenchymal cell shape (Hay, 1973).

The formation of rosette-like, cystic and tubular epithelial structures within theamorphous cell mass apparently occurs during normal morphogenesis of renaltubules and is also comparable with the formation of cavities in aggregates frommixed suspensions of amphibian embryonic cells (Townes & Holtfreter, 1955).Epithelial structures in the posterior end of the mouse embryo such as thesecondary neural tube (Criley, 1969; Schoenwolf, 1984) and the tail gut (Svajger,


Kostovic-Knezevic, Bradamante & Wrischer, 1985) develop by an analogousmechanism. Cystic and tubular transformation of the mesenchyme also occursin mouse embryos bearing T-locus mutations (Spiegelman & Bennett, 1974;Spiegelman, 1976; Fujimoto & Yanagisawa, 1979).

COMPOSITION OF TERATOMAS

Course of development

In a previous study the course of development of isolated head-fold-stageembryonic ectoderm as a renal isograft was recorded at 2- to 3-day intervals(Levak-Svajger & Svajger, 1979). The above mentioned breaking up into mesen-chyme and the formation of epithelial cysts took place during the first three daysafter transfer. After 5 days the graft enlarged and became vascularized. Itcontained large amounts of immature neural tissue with areas of massive cellnecrosis. The initial formation of the epidermis and myotubes could be observed.The rest of the mesenchyme was undifferentiated. After 7 days the differentiationof the neural tissue and of the epidermis was advanced and the first chondrogenicblastemas appeared. Nine days after transfer the grafts contained abundant neuraltissue, striated muscle and cartilage. Ossification started and the epidermis showedthe onset of keratinization and of the formation of hair buds. Adipose tissueappeared after 12 days. There was a considerable variation in the size of tumours,even within a single experimental series. In general, it paralleled the duration ofdevelopment in the host. There was a consistent transition from a discoid shapeto massive, tuberous teratomas which only rarely penetrated into the renalparenchyma.

Tissue differentiation

The chaotic arrangement of mature tissues, often associated in a clearlyrecognizable organotypic combination (Skreb & Svajger, 1975; Svajger, Levak-Svajger, Kostovic-Knezevic & Bradamante, 1981) conformed to the definition andcharacterization of benign or mature embryo-derived teratomas (Damjanov &Solter, 1974; Solter, Damjanov & Koprowski, 1975) and of spontaneous benignteratomas observed in man (Willis, 1935,1962; O'Hare, 1978) and in some strainsof mice (Stevens & Hummel, 1957; Stevens, 1967). The tissue composition ofteratomas varied according to the initial developmental stage of the ectoderm (seethe next section), but in the most favourable cases (the primitive ectoderm graftedfor 30 days) the tumours regularly consisted of tissue derivatives of all threedefinitive germ layers in the classical sense of the germ layer definition:

'Ectodermal tissues': skin (epidermis, hairs, sebaceous and mammary glands),neural tissue (brain, neural retina, choroid plexus, ganglia) and others (lentoids,oral cavity, teeth, salivary glands).

'Endodermal tissues': foregut-derived epithelia (pharynx, tongue, salivaryglands, thyroid, thymus, parathyroid, oesophagus, stomach, respiratory tube,


lung, pancreas), midgut- and hindgut-derived epithelia (small and large intestine,urogenital sinus, prostatic gland).

'Mesodermal tissues': white and brown adipose tissues, cartilage, membraneand endochondral bone, smooth, skeletal and cardiac muscle.

Well-expressed organotypic differentiation and combinations of tissues werecomprehensively described and discussed in our previous publications (Skreb &Svajger, 1975; Svajger, Levak-Svajger, Kostovic-Knezevic & Bradamante, 1981).The endoderm-derived epithelia were outstanding in displaying a wide range ofsegment-specific features. In general, characteristic associations of tissues, whichnormally occur during embryogenesis in situ, also occurred in teratomas: pseudo-stratified ciliated columnar epithelium-I-hyaline cartilage (respiratory tube),stomach and intestinal epithelium with glands + layers of smooth muscle withintramural ganglia (eosinophils and plasmocytes in the lamina propria!), epen-dyma of the choroid plexus + an ample subependymal vascular network.

With the occasional exception of the neural retina, the mature neural tissue,which histologically mimicked the brain, did not display the cytoarchitectonicstypical of any particular segment of the central nervous system. According toobservations on the neurulating mouse embryo in situ, the anterior part of theembryonic ectoderm (rostral to the Hensen's node) gives rise to all segments of thebrain (Tamarin & Boyde, 1976; Jacobson & Tam, 1982; Meier & Tarn, 1982).There are also structures which never differentiate in teratomas (liver, kidney,adrenal, gonads, germ cells, pituitary, notochord). This could probably beexplained by the lack of the delicate topographically and chronologically regulatedtissue displacements and interactions which operate during the normal develop-ment of these organs in situ. The primordial germ cells are first distinguishable atthe posterior end of the late-primitive-streak mouse embryo (Heath, 1978; Tam &Snow, 1981; Snow, 1981; McLaren, 1983; Beddington, 1983a). Their positionbefore and early in gastrulation is not known and very probably they were notincluded in the primitive ectoderm isolated for grafting in our experiments.

Incidence of differentiated tissues

The grafting experiments with isolated germ layers have been based on thepresumption that the tissue composition of resulting teratomas reflected the initialdevelopmental potential of the isolated cell population. However, this potentialcannot be evaluated only in terms of the incidence of tissues differentiated ingrafts. The major characteristics of the experimental embryonic teratomas is theirstructural polymorphism. The presence of different tissues varies a great dealamong grafts of the same embryonic origin (Skreb, Svajger & Levak-Svajger,1971; Levak-Svajger & Svajger, 1971,1974). The tissue composition of teratomasis therefore most probably dependent on more factors than the mere develop-mental capacity of the initial group of embryonic cells. The following factors mightbe involved, (a) Massive cell necrosis, which exceeds the normal rate of earlyectodermal cell death in situ (Poelmann & Vermeij-Keers, 1976) occurs within theimmature neural tissue a few days after transfer (Levak-Svajger & Svajger, 1979)


as well as in the mature neural tissue later on (Svajger, Levak-Svajger, Kostovic-Knezevic & Skreb, 1985). At this time the neuroectoderm grows intensively andthis growth is not supported by adequate vascularization. In grafts in which thedelay of blood supply is not too long, the developing neural tissue can recoverfrom the initial cell loss and subsequently differentiates into masses of matureneural tissue (the usual feature in large tumours). On the other hand, in sometumours, very probably because of prolonged inadequate vascularization, theneural tissue undergoes further regression and sometimes almost completelydisappears. That the amount and quality of tissue differentiation is dependent onan effective graft incorporation and vascularization became evident in exper-iments with intraocular grafts of the rat egg-cylinders (Skreb & Levak, 1960;Levak-Svajger & Skreb, 1965). (b) Tissues which originate from large presumptive(prospective) areas (neuroectoderm, gut, mesenchymal tissues) differentiate morefrequently and in larger amounts than those which in situ develop from restrictedareas of the germ layers (teeth, stomach, thymus, thyroid, parathyroid), (c) Sometissues need more time for their final differentiation and therefore appear morefrequently in tumours after 30 days (adipose tissue, bone, skeletal muscle, deriva-tives of the foregut etc.). (d) The phenotypic expression of 'endodermal' differen-tiation might be masked by the accumulation of intracystic fluid. The resultingdistended cysts do not display organotypic epithelial features but are lined by anon-specific, flattened epithelium (Usadel, Rockert, Obert & Schoffling, 1970).

Summing up, the final phenotype expressed in renal isografts of early ratembryonic germ layers results not only from their initial developmental capacities,but also from various, usually ill-defined and uncontrolled events which occurwithin the interval between the tissue isolation and the fixation of teratomas forhistology.

GERM LAYER ORIGIN OF TISSUES

The classical view of germ layer origin of various tissues has been modified bythe experience that the same tissues can originate from different definitive germlayers (see De Beer, 1947). It is therefore hardly possible to interpret with absolutecertainty the complex structure of teratomas in terms of the provenance ofdifferent tissues from particular germ layers. This holds especially for sometypically 'mesodermal' tissues which may develop from the mesenchyme of bothmesodermal and ectodermal (neural crest, tail bud) origin. In this review we usethe conventional terms for the germ layers and of their tissue and organderivatives, while constantly keeping in mind the very restricted developmentalvalue of these terminological designations.

Neural tissue and epidermis are exclusively of ectodermal origin. Neuroecto-derm is the first tissue which appears and grows in grafts of the primitive ectoderm(Levak-Svajger & Svajger, 1979). Epidermis develops from undifferentiated epi-thelial cysts, which form early after grafting (Svajger et al. 1981). The finalepidermal phenotype is evident when hair buds appear on the basal side of the


keratinized squamous epithelium. The same type of epithelium (without the skinappendages) may be of endodermal origin. It lines the mucous membranes of theoesophagus and the cardia of the stomach.

Ciliated columnar, pseudostratified epithelium with goblet cells and adjacentglands is the most common epithelial structure in teratomas. It is usuallyconsidered as an endodermal tissue, but the nasal cavity which is of ectodermalorigin, is also lined by this type of epithelium. Even the large salivary glands seemto be of dual germ-layer origin. The parotid gland is probably of ectodermalorigin, while the submandibular and sublingual glands are endodermal derivatives(Hamilton, Boyd & Mossman, 1976). This view is supported by the experiencethat the parotid gland acts as an integumental derivative and the submandibulargland as a gut derivative in tissue recombination experiments (Kollar, 1983).

The epithelial lining of the intestine is nowadays commonly considered as aderivative of the primitive ectoderm. Not only the experimental evidence (seeRossant & Papaioannou, 1977 and Beddington, 1983£ for a review), but alsomorphological observations indicate this cell lineage (Jolly & Ferester-Tadie,1936; Jurand, 1974; Poelmann, 19816). A minor contribution of the primitiveendoderm to the hindgut was suggested by Jolly & Ferester-Tadie (1936). In arecent publication Jackson et al (1981) argued on the basis of ultrastructuralfeatures that the ileum, unlike the rest of the small intestine, is derived from theprimitive embryonic endoderm. However the ultrastructure of a cell type is notnecessarily dependent on its germ layer origin. As an example, all epithelial cellsinvolved in the transport of water and ions display the same general ultrastructuralfeatures, regardless of their embryonic origin (choroid plexus, ciliary bodyepithelium, stria vascularis of the inner ear, striated ducts of the large salivaryglands, convoluted tubules of the kidney). On the other hand, in all our ex-periments with the primitive ectoderm as a renal isograft, all segments of theprimitive gut developed (including the very posterior hindgut derivatives: the largebowel and the urogenital sinus). It seems improbable that only a short segment ofthe small intestine would have a different origin.

Epithelial organs of mesodermal origin (kidney, adrenal cortex, gonads) neverdeveloped in grafts. Mesenchymal tissues, however, appeared regularly and madeup a considerable part of teratomas. Except for adipose tissue and cardiac muscle,mesenchymal tissues could also have developed from neural crest cells (Le Lievre& Le Douarin, 1975; Nakamura, 1982; Nakamura & Ayer-Le Lievre, 1982). Invarious experimental systems and in some pathological conditions in situ skeletalmuscle has been described as developing from the thymus reticulum (Wekerle,Paterson, Ketelsen & Feldman, 1975; Ketelsen & Wekerle, 1976), Schwann cells(Wallace, 1972; Maden, 1977), pituitary cells (Brunner & Tschank, 1982), glialcells (Lennon, Peterson & Schubert, 1979) and different parts of the centralnervous system (for a review see Inestrosa, 1982).

The thymus is perhaps the most intriguing example of unexpected tissuepotentialities and differentiation. It develops as an epithelial derivative of the thirdpharyngeal endodermal pouches, which acquire a mesenchymal (or reticular)


appearance after the invasion of lymphoid stem cells. In the human thymusprimordium during the 8th gestational week as well as in the adult human thymus,myoid (myoepithelial) cells have been found, which displayed some morpho-logical features of smooth muscle cells, but showed immunoreactivity with theantisera prepared against striated-muscle-type myosin (Puchtler, Meloan, Brauch& Gropp, 1975; Drenckhahn, Unsicker, Griesser, Schumacher & Groschel-Stewart, 1978; Gaudecker & Miiller-Hermelink, 1980).

One may conclude that, although skeletal muscle is a typical 'mesodermaPtissue, all three definitive germ layers bear at least a latent myogenic potential. Onthe other hand, myoblasts can differentiate into hyaline cartilage when grown onan appropriate substratum (Nathanson & Hay, 1980).

In addition to cardiac muscle, brown adipose tissue seems to be the most typicalderivative of the mesoderm. This tissue was the only one to develop in renal graftsof isolated rat embryonic mesoderm (Skreb, Svajger & Levak-Svajger, 1976).

The tail bud, which in all classes of vertebrates appears behind the posteriorneuropore, represents a mass of mesenchyme with an extraordinary origin andfate. It forms from the remnants of the primitive streak and by the partialdissociation and migration of neuroectodermal cells of the terminal neural tube. Itdifferentiates into caudal somites and their derivatives, and even back into theneuroepithelium (the 'secondary neurulation' and, or, the 'secondary body forma-tion', Holmdahl, 1935; Bijtel, 1936; Jolly & Ferester-Tadie, 1936; Peter, 1941,1951; Tarn, 1981; Tarn, Meier & Jacobson, 1982; Nakao & Ishizawa, 1984;Schoenwolf, 1984; Schoenwolf & Nichols, 1984). Very probably it also contributescells to the tail gut (Butcher, 1929; Svajger etal. 1985). Thus the tail portions of theneural tube and of the gut develop by aggregation from the diffuse cell mass of thetail bud and their development is strongly reminiscent of the morphogenesis anddifferentiation of the embryonic ectoderm in renal grafts. It is evident that thecapacity of the ectoderm to produce cells with a 'mesodermal' fate (the formationof mesenchymal tissues from the neural crest and the tail bud), extends beyond theperiod of gastrulation. This problem will be further considered in the next section.The data reviewed in this section confirm De Beer's (1947) view that one and thesame tissue type can originate from different definitive germ layers.

RESTRICTION OF DEVELOPMENTAL CAPACITIES OF THE EMBRYONIC ECTODERM

DURING GASTRULATION

In the rodent embryo the primitive embryonic ectoderm appears as a distinctcell mass during the blastocyst stage. Its basal Surface is lined by the primitiveendoderm. During the time interval between the blastocyst stage and the onsetof gastrulation, both these primitive (primary) germ layers as a whole (thetwo-layered embryonic shield) undergo remarkable changes in their geometry(inversion of germ layers), which lead to the formation of the so-called egg-cylinder (see Rossant & Papaioannou, 1977 for review). This unique shape ofpregastrulation (preprimitive streak) rat and mouse embryos creates severe


difficulties in both the microsurgical manipulation of the embryo and theinterpretation of the isolation and grafting experiments. The latter concernsespecially the construction of regionally well-defined fate maps.

We have experimentally approached the problem of diversification of celllineages during gastrulation by grafting the isolated rat embryo germ layers,excised areas of them and particular germ layer combinations under the kidneycapsule of adult male syngeneic animals, and by recording the outcome ofdifferentiation and morphogenesis by the histological analysis of the resultingteratomas. In the following text we will refer to the experimental series designatedI-XIV in Table 2.

The embryonic ectoderm as a whole

The primitive embryonic ectoderm of the preprimitive streak stage gives riseto tissue derivatives of all three definitive germ layers: neural tissue, skin,'mesodermal' tissues and derivatives of the primitive gut (series I, Levak-Svajger& Svajger, 1971).

Embryonic ectoderm isolated together with the initial mesodermal wings at theearly primitive streak stage also differentiates into tissue derivatives of all threedefinitive germ layers (series V, Levak-Svajger & Svajger, 1974).

At the head-fold stage the isolated embryonic ectoderm (regardless of thepresence or absence of the primitive streak) differentiates only into 'ectodermal'and 'mesodermal' tissues (series IX, Levak-Svajger & Svajger, 1974).

At all these developmental stages the isolated embryonic endoderm shows nocapacity for differentiation as an isograft but becomes completely resorbed shortlyafter transfer (series IV, VIII, XII, Levak-Svajger & Svajger, 1974).

When at the head-fold stage the definitive endoderm is isolated and graftedtogether with the adjacent mesoderm, it differentiates into a variety of derivativesof the primitive gut (series X, Levak-Svajger & Svajger, 1974).

At the same developmental stage the isolated embryonic mesoderm differen-tiates only into the brown adipose tissue (Skreb et al. 1976). However, incombination with the definitive endoderm, it develops into a variety of typical'mesodermal' tissues (adipose tissue, cartilage, bone, muscle, series X, Levak-Svajger & Svajger, 1974).

These results, confirmed by similar findings in testicular grafts of the mouseprimitive ectoderm and endoderm (Diwan & Stevens, 1976) indicate that prior togastrulation the progenitor cells of all the tissues of the foetal body are locatedwithin the primitive embryonic ectoderm. During gastrulation first the endo-dermal and then the mesodermal cells leave the primitive ectoderm and becomedisplaced to the topographic positions of the respective definitive germ layers.What remains in the topographic position of the primitive ectoderm, becomes thedefinitive ectoderm of the three-layered embryonic shield (the late primitivestreak or the late gastrula stage) which subsequently (at the head-fold stage)undergoes neurulation (see Rossant & Papaioannou, 1977; Beddington, 1983b).From this stage on these three cell sheets have been considered as the definitive

11 13

Tab

le 2

. T

ypes

of

rena

l iso

graf

ts r

evie

wed

in th

is p

aper

Stag

e(W

itsch

i)*

Stag

ede

sign

atio

n

Rat

em

bryo

age

(day

s)

Cor

resp

ondi

ngag

e of

the

mou

se e

mbr

yo(d

ays)

Exp

erim

enta

l se

ries

(Typ

e of

gra

ft)

Prep

rim

itive

str

eak

(pre

gast

rula

)

12

Ear

ly p

rim

itive

str

eak

(ear

ly g

astr

ula)

Hea

d fo

ld(l

ate

gast

rula

, ea

rly

neur

ula)

6 I.

W

hole

pri

m,

ecto

derm

II.

Tra

nsve

rsal

hal

ves

of t

he p

rim

, ec

tode

rmII

I.

Lon

gitu

dina

l ha

lves

of

the

prim

, ec

tode

rmIV

. Pr

im,

endo

derm

6-5

V.

Who

le p

rim

, ec

tode

rm +

mes

oder

mal

win

gsV

I.

Tra

nsve

rsal

hal

ves

of t

he p

rim

, ec

tode

rm +

mes

oder

mal

win

gsV

II.

Lon

gitu

dina

l hal

ves o

f the

pri

m. e

ctod

erm

+ m

esod

erm

alw

ings

VII

I.

Prim

, en

dode

rm

7-5

IX.

Ect

oder

m (

with

or

with

out

the

prim

itive

str

eak

regi

on)

X.

End

oder

m +

mes

oder

mX

I.

Mes

oder

mX

II.

End

oder

mX

III.

A

nt.

post

, an

d di

stal

par

ts o

f th

e eg

g-cy

linde

rX

IV.

Axi

al a

nd l

ater

al s

trip

s of

the

ect

oder

m

I

See

New

(19

66).


germ layers in the classical topographic, histogenetic and organogenetic sense(Oppenheimer, 1940).

The essential conclusion that all the foetal primordia are descended from theprimitive embryonic ectoderm was anticipated in the pioneering experiments ofGrobstein (1952) who showed that the early mouse embryonic ectoderm retainedits capacity to differentiate into the gut epithelium even as an intraocular graftafter 5-7 days of preculturing in vitro. This is also consistent with (1) themorphological observations of the continuity of the Hensen's-node-derived cellswith the epithelium of the definitive gut (Jolly & Ferester-Tadie, 1936; Jurand,1974; Poelmann, 1981a,b), and (2) the results obtained by injecting single 5th daymouse primitive ectoderm cells into the 4th day blastocyst. The progeny of theinjected cells contributed to tissues and organs throughout the foetus, regardlessof their germ layer origin (Gardner & Rossant, 1976, 1979). The rodent embryoshares this basic principle of definitive germ layer formation during gastrulationwith the avian embryo, in which it was studied by similar or more direct methods(Nicolet, 1971; Veini & Hara, 1975; Fontaine & Le Douarin, 1977). The analogyof the general developmental principles in these two classes of vertebrates isfurther suggested by the experimental findings that the embryonic ectoderm of therabbit is competent to respond to the primary organizer of the chick embryo(Waddington, 1936), and that a secondary neural groove can be induced in the ratembryonic ectoderm by implanted medullary plate material (Toro, 1938; seeWaddington, 1957 for a review).

The gradual restriction of developmental capacities of the initially pluripotentprimitive embryonic ectoderm and the gradual 'release' of some of these capacitiesto the newly formed endoderm and mesoderm conform with the claim that "it isonly legitimate to speak of the separate germ layers when their segregation fromone another is complete" (De Beer, 1958). The segregation of definitive germlayers coincides with the end of gastrulation. However, during the subsequentneurulation, and even later, the cells of the neural crest and of the tail budoriginate from the neural ectoderm and differentiate into typical 'mesodermal' and'endodermal' tissues (see the section entitled Germ Layer Origin of Tissues).Moreover the 'mesectoderm' or 'ectomesenchyme' of the tail bud gives rise to thecaudal portion of the neural tube, of the neural crest and of the tail gut(Schoenwolf, 1984; Schoenwolf & Nichols, 1974; Svajger et al 1985). Thiscorroborates the already old view of the dual origin of the vertebrate body: theprimary or indirect development of anterior structures of the body during whichthe formation of three germ layers precedes the development of the embryonicorgan rudiments, and the secondary or direct development of the posteriorstructures of the body without the intervening formation of germ layers(Holmdahl, 1935; Peter, 1941,1951).

We may conclude that the embryonic ectoderm never exists as a discrete,definitive germ layer in the classical sense of the word (i.e. as the germ layer withcapacities restricted to the formation of the neural tissue and the epidermis). Inother words, the three-layered embryonic shield directly transforms only into the


anterior portion of the foetal body. These conclusions are in strict contradictionwith the rationale of the classical germ layer theory.

Recently, Lovell-Badge, Evans & Bellairs (1985) investigated the proteinsynthetic patterns of tissues in the chick embryos from gastrulation and early post-gastrulation stages. This study revealed polypeptides which may be considered asmarkers for endodermal and mesodermal cell types, but no polypeptides uniquefor the ectoderm were found. The lack of biosynthetic specialization is in apparentagreement with the capacity of the (neuro)ectoderm to originate various types oftissues even at postgastrulation stages.

Regionalization of the ectoderm and the fate map problem

At both the preprimitive streak and the early primitive streak stages the halvesof longitudinally or transversely cut egg-cylinders or isolated primitive ectodermsgive rise to tissue derivatives of all three definitive germ layers and there is noindication of a regionally restricted tissue- or organ-forming capacity of theprimitive ectoderm (series II, III, VI, VII, Svajger etal. 1981).

At the head-fold stage the ectoderm displays some regionally specific de-velopmental capacities: ectodermal structures of the head (choroid plexus, teeth,vibrissae) develop only from the anterior part of the egg-cylinder (series XIII,Levak-Svajger & Svajger, 1974).

At the head-fold stage the axial and the lateral strips of the ectoderm do notdiffer in their capacities to differentiate into both neural tissue and epidermis andits derivatives (series XIV, Svajger & Levak-Svajger, 1976).

Attempts to construct on the embryonic ectoderm a map of regions withrestricted developmental fates and, or, potencies are relevant to one of the majorproblems in developmental biology: the mosaic- versus regulative-developmentdilemma. Recently this has been subject to various experimental approaches in theearly postimplantation mouse embryo.

Beddington (1981) introduced the technique of injecting clumps of about 20[3H]thymidine-labelled embryonic ectoderm cells into the embryonic ectoderm ofunlabelled synchronous mouse embryos. The labelled cells were detected inradioautographs of histological sections of host embryos after a 36 h period ofcultivation in vitro. Using this method Beddington (1982) carried out a detailedanalysis of the developmental fate of embryonic ectoderm cells of the 8th daymouse embryo (late primitive streak). She isolated the donor cells from theanterior, distal and posterior areas of the egg-cylinder and injected themorthotopically and heterotopically into these areas of synchronous hosts. Thedistribution of injected cells was recorded when the host embryos had attainedin vitro the early somite stage. In orthotopic injections the results were as follows:(a) the anterior ectoderm gave rise to predominantly neuroectoderm and surfaceectoderm; (b) the distal ectoderm (the area of the Hensen's node) generateddefinitive gut endoderm, notochord and embryonic mesoderm; (c) the posteriorectoderm (the primitive streak) contributed to the embryonic and extraembryonicmesoderm. In the heterotopic series of injections the distal and posterior


embryonic ectoderm conformed to the colonization patterns characteristic oftheir new location and their potency always exceeded their developmental fatein situ. The anterior embryonic ectoderm preferentially differentiated into typicalectodermal structures. In all three regions the ectoderm showed the capacity toform 'mesodermal' derivatives. This pluripotency of ectodermal cells conformsto the evidence from other experimental systems, that "the isolated part tends toform a greater variety of structures than when left in place in the embryo"(Harrison, 1933). These results speak against the existence of a rigid mosaicism inembryonic ectoderm, which would be resistant to regulative influences fromsurrounding cells. So do our results with the grafting of isolated parts of pre-primitive streak and early primitive streak rat embryonic ectoderm (Svajgeret al. 1981). It ought to be stressed, however, that the egg-cylinders were tran-sected by randomly oriented cuts only into longitudinal or transverse halves. Evena more elaborate design of isolation of small areas could hardly overcome thedifficulties caused by the complex topology of the inverted rodent embryonicshield.

Notwithstanding the above mentioned problems, Snow (1981) isolated circum-scribed areas of the whole mouse egg-cylinder (all three germ layers) at the early-,mid- and late-primitive-streak stages, and recorded their development as well asthe development of the rest of the mutilated cylinder after a 24 h period ofcultivation in vitro. He could demonstrate an autonomous development of isolatedpieces and an absence of regeneration or regulation in the rest of the embryonicshield. He interpreted these results as suggesting a mosaic or polyclonal type ofdevelopment of the early mouse embryo. It is however dubious whether in thetopologically enigmatic geometry of the early egg-cylinder one would be able tostrictly delineate areas with mutually exclusive and particular developmental fateswithin any of the constituent germ layers. These operations might also disturb thechanges in the relative positions of ectoderm and mesoderm brought about by thedifferential growth pattern of the embryonic ectoderm (Poelmann, 1980a, b,1981a, fr; Tarn & Meier, 1982). The current knowledge on the nature and specificityof epithelial-mesenchymal interactions in development does not yield sufficientinformation for a convincing interpretation of this experimental system (Lawson,1974; Yasugi & Mizuno, 1978; Ishizu^a-Oka & Mizuno, 1984). As pointed out byBeddington (1982), the fact that the fragments isolated by Snow were composed ofall three germ layers, does not allow us to draw any conclusions concerning themosaic nature of the embryonic ectoderm alone.

Although Beddington's (1981, 1982) results with embryonic chimaeras are ingeneral agreement with the results of our grafting experiments (Levak-Svajger& Svajger, 1974; Svajger et al. 1981), some differences in the design of theexperiments must not be neglected. The grafting of isolated germ layers, wholeegg-cylinders or their dissected parts differs from the injection of small groups ofcells with regard to the mass and coherence of the ensemble of transferred cells. Inthe in vitro experiments the results are recorded after 24-36 h, when the cells inquestion have neither yet reached their final positions nor displayed all their


developmental potencies. On the other hand, in grafting experiments a final levelof differentiation is attained, but the cell allocation pattern cannot be studiedbecause of the disorganized growth and morphogenesis within the graft.

More recently Beddington (1983a) isolated anterior and posterior slices of boththe whole 8th day mouse egg-cylinder and the isolated ectoderm. There were nosignificant differences between the histogenetic potentials of these fragments intesticular grafts. Both of them generated endodermal structures (epithelium of theprimitive gut). This result, as well as the differentiation of the gut epithelium fromthe orthotopically injected distal ectodermal cells (Beddington, 1981,1982), couldmost probably be explained by the relatively early developmental stage of theembryonic shield at the time of manipulation (late primitive streak). At this stage(before the head-fold formation) the embryonic ectoderm has probably not yetexhausted all its prospective endodermal and mesodermal cells, and in the area ofHensen's node the cells have still retained the tendency to invaginate into theexternal endoderm layer. At the subsequent head-fold stage gastrulation iscompleted and the embryonic ectoderm no longer contains prospective endo-dermal cells (Levak-Svajger & Svajger, 1974). In combination with the mesodermboth the ectoderm and the endoderm display some regionally specific develop-mental capacities (Svajger & Levak-Svajger, 1974). Whether or not at this stagethe ectoderm alone is still capable of forming true mesoderm in renal grafts isuncertain and probably represents a mere semantic question of limited explana-tory value in the light of data concerning the developmental capacities of theneural crest and the tail bud (see previous sections).

One might assume that with regard to the overall developmental capacity of theectoderm and to its regionally restricted potential Beddington's (1981, 1982,1983a) experiments in the mouse concern the short developmental intervalbetween the early-primitive-streak stage and the head-fold stage which weinvestigated in the rat embryo (Levak-Svajger & Svajger, 1974; Svajger &Levak-Svajger, 1974). At the same developmental stage in the mouse embryo,the anterior ectodermal cells injected with horseradish peroxidase contributeprimarily to the brain (Lawson, Meneses & Pedersen, 1983).

Tarn (1984) transplanted the posterior portion of the primitive-streak-stage (7-5days) mouse egg-cylinder beneath the kidney capsule. He obtained teratomascomposed of tissue derivatives of all three germ layers, but only the presence ofhindgut and mesodermal urogenital structures suggested regionally specificdifferentiation.

The topology and the chronology of cell commitment within the definitive ratembryonic ectoderm is now being subject to analysis in our laboratory. The factthat both the axial and the lateral strips of the head-fold stage ectoderm bear thecapacity for neuroectodermal and epidermal differentiation (Svajger & Levak-Svajger, 1976) has to be interpreted with respect to the recent evidence of a strictdemarcation of these two ectodermal cell lineages with regard to their cell surfaceglycoconjugates (Currie, Maylie-Pfenninger & Pfenninger, 1984; Thiery et al.1984).


To sum up, the results of our and Beddington's experiments apparently confirmor, at least, do not contradict the classical experimental evidence that in severalclasses of the chordate phylum the isolated parts of the early pregastrulation orprimitive ectoderm (the blastula ectoderm) are able to regulate to a considerableextent. A restriction of the developmental potential appears during gastrulationand becomes more pronounced (regionalization) at subsequent stages (seeWaddington, 1957 and Snow, 1981 for discussion). A fate map comparable to thatconstructed for the amphibian embryo would only have an explanatory value if itwere possible to sharply delineate embryonic cell populations with distinct de-velopmental fates while they are still resident within the same, initial germ layer(the primitive ectoderm). When gastrulation starts the displacement of cell groupsbrings about new spatial relationships, interactive events and cell commitmentoccur and are followed by increasing regional restrictions of developmentalpotencies. A fate or allocation map constructed at the gastrula (primitive streak)stage, as in Snow's (1981) experiment, can therefore give little or no information infavour or against the mosaic or clonal nature of the early postimplantationdevelopment of the rodent embryo.

That even a precisely constructed fate map of prospective (presumptive) areason the pregastrulation (primitive) ectoderm can fail to reflect the true de-velopmental repertoire of circumscribed groups of embryonic cells, can best berevealed by a careful analysis of data obtained by mapping the amphibianblastoderm. These data represent the basic common knowledge in all textbooks ofembryology for more than half a century. The classical schema of the de-velopmental mosaic of the amphibian blastula and early gastrula was constructedon the basis of Vogt's (1929) vital surface staining and the isolation experiments ofHoltfreter (1938). The general conclusion has been that all the future organrudiments and tissue derivatives of definitive germ layers are represented bycircumscribed cell groups on the surface of the late blastula and early gastrulawall. According to Holtfreter, some of these primordia have the capacity forautonomous development (autodifferentiation) when isolated and explantedin vitro. However, more recent studies on amphibian gastrulation have revealeda number of details which do not conform to the classical view, (a) The epi-thelial wall of the blastula and early gastrula is a heterogeneous cell populationdistributed in superficial and deep positions; (b) these cell groups are subject toindependent shifting and displacement during gastrulation (Keller, 1980; L0vtrup,1983); (c) the cells in different layers of the blastula wall differ in severalproperties, including prospective significance, inasmuch as "the differences in theproperties of cells from the primitive ectoderm inner cell layer in various pre-sumptive regions are less than between the cells of the outer and inner layers in thesame presumptive region" (Dettlaff, 1983); (d) when small groups of 15-20 cellsfrom the late blastula wall are isolated and explanted in vitro, only the superficialcells form small epidermal vesicles while the deep cells do not differentiate at all(L0vtrup, 1974,1975,1983; Landstrom & L0vtrup, 1979).


It may be concluded from these experimental results that by isolating the'ectodermal' cells of the amphibian blastula and early gastrula the authors haveactually isolated cells belonging to at least two definitive germ layers (Nieuwkoop,1973; L0vtrup, 1975). Only the ectoderm, the notochord and the endoderm appearin the form of presumptive areas on the surface of the blastula. The prospectivemesodermal cells are never present on the surface of the pregastrulation embryo.The 'autodifferentiation' of prospective mesodermal cells from the surfacematerial occurs as a consequence of interaction of embryonic cells with differentprospective germ layer fates. It is therefore interesting to notice that in a mostrecent publication the early amphibian embryonic cells were defined on the basisof their position in different regions of the gastrula rather than of their germ layerorigin (Cleine & Slack, 1985).

Coming back to the problem of terminology, it seems that the German, Frenchand Italian terms for germ layers (Kleimblatter, feuillets embryonnaires, fogliatigerminativi = germ leaves) are more appropriate than the English term germlayers, because they denote stretched or extended cell populations withoutprejudging their monolayer nature.

It has been believed that the amphibian embryo differs from the mammalianone by the early determination of bilateral symmetry (grey crescent), the earlycommitment of its cells to different developmental pathways if isolated at theblastula stage (Nieuwkoop, 1973) and by the corresponding early susceptibility toexogeneous teratogenic influences (Rugh, 1954). However, it was recently claimedthat the vegetal blastomeres of X. laevis embryos implanted to another blastula didnot become determined before the beginning of gastrulation. At the midblastulastage they contribute progeny to all three definitive germ layers (Heasman, Wylie,Hausen & Smith, 1984). Contrary to previous data from isolation experimentsthese results suggest that also in the amphibian embryo, at least in some of itsregions, the restriction of the developmental potential of the primitive ectodermmay start not earlier than at the early gastrula stage.

CONCLUDING REMARKS

The experimental results with ectopic grafts reviewed here apparently reconfirmthe view on the common origin of all cells of the foetal body from the primitiveembryonic ectoderm, and a gradual topographical and developmental diversifi-cation of cell lineages during gastrulation. However, the methodology of histo-logical analysis of experimental teratomas imposes a number of difficulties andlimitations in the interpretation of findings in dynamic developmental terms.Besides the fact that the final tissue composition of tumours is largely dependenton the "survival of fittest cells in conditions that are never quite ideal"(Waymouth, 1974), the peculiar morphogenetic behaviour of the primitive ecto-derm as a renal graft is quite different from the ordered cell displacement in theliving early embryo. Some other problems were stressed before in the classicalstudy of Grobstein (1952): "It seems important to keep in mind that from the (...)


differentiative behaviour of such plastic systems it is exceedingly difficult to drawconclusions about the state of determination of their component cells at the time ofisolation. In any situation where opportunity is offered for reorganization -particularly the case in transplant, explant and defect experiments - regulationmay lead to an increase of cell types or simplification of organization to a decreaseof cell types, relative to the prospective significance of the area in question. Whatis in fact tested in such experiments is the state of determination of the system, notof the component cells."

The non-specificity of the classical germ layers as to their tissue-forming capacityloads the interpretation with additional problems and demands for a more flexiblethinking about cell lineages in early development. The dual origin of most ofthe 'mesodermal' tissues is perhaps the best example. The mesenchyme with thepotency to form supportive tissues and muscle originates from three sources: theprimitive streak, the neural crest and the caudal neuroectoderm (the tail bud)."There is, however, no essential differences between them, because they originatedirectly or indirectly from the same ectoderm, and the classical distinction seemsto be reducible to the mere difference in stage of mesodermal formation"(Nieuwkoop, 1973). Once again it becomes clear that the germ-layer theory is amorphological concept, and has nothing to do with developmental potencies ofembryonic tissues (De Beer, 1947). It is noteworthy that long ago Kolliker (1879,quoted in Oppenheimer, 1940) claimed that "all three germ layers possess thepotency and the capacity also for transformation into all tissues, but because oftheir specific morphological configurations they cannot everywhere manifest thispower".

We may summarize that from the histological analysis of teratomas one can-not draw conclusions on the germ layer origin of tissues but only about thedevelopmental potency of topographically and chronologically defined, isolatedregions of the embryo. This seems to restrict the usefulness of the method of ectopictransfer of early mammalian embryonic tissues. However, various experimentalapproaches on mammalian and avian embryos have given analogous orcomplementary results which permit us to believe in the existence of commonprinciples of cell lineages in the living embryo and in experimental transplants.

We end with a list of some questions which have arisen from our graftingexperiments but cannot yet be answered.

(a) Do particular tissues develop in teratomas from the same precursor cells andby interaction with the same associated cells as they would have in normaldevelopment?

(b) How is the neural tissue induced in the absence of a morphologically definedprimary organizer? Can we suppose the formation in grafts of an atypical inductivecentre, which is able to induce the differentiation of neural tissue but not toorganize it morphogenetically?

(c) Do local cell-cell or cell-substrate interactions occur in teratomas, and, ifso, do they influence differentiation more than cell lineage, as demonstrated forthe mouse extraembryonic endoderm (Hogan & Tilly, 1981)?


(d) Do the transdifferentiation-like phenomena (Yamada, 1977) play any role inthe histogenesis within teratomas?

(e) Is the epithelial-mesenchymal conversion of ectodermal grafts a solelymorphological phenomenon, or is it correlated with a reorganization of thecellular activity, such as a switching in the expression of cytofilaments fromcytokeratins to vimentin (Franke etal. 1982)?

The problem in the broadest sense might perhaps best be revealed byacknowledging the wisdom of one of the old masters of the discipline: "Thereis always room for fallacy, even when the logical procedure may seem unim-peachable, and no conclusion in embryology is safe if based upon but a singleproof. This, to some, may all seem purely formal and of no practical consequence.It is, nevertheless, important to realize that even the language of science is stillbound by tradition (...) and it is by no means free from anthropomorphism andrelicts of our demonology, which are difficult to escape and which may not onlylend a false sense of security to our explanations but also may suggest foolishquestions that never can be answered" (Harrison, 1933).

The original work was supported by grants of the Self-managing Community of Interest inScience of S. R. Croatia (SIZ-V), through funds made available to the U.S.-Yugoslav JointBoard on Scientific and Technological Cooperation (No. 02-094-N) and by Small SuppliesProgram of the World Health Organization, Geneva, Switzerland. The authors are grateful toDr R. L. Gardner and to Dr R. Beddington for their criticism and helpful suggestions during thepreparation of the manuscript.

REFERENCES

ADAMSON, E. D. & AYERS, S. E. (1979). The localization and synthesis of some collagen types indeveloping mouse embryo. Cell 16, 953-965.

ARIENS KAPPERS, J. (1941). Kopfplakoden bei Wirbelitieren. Ergebn. Anat. EntwGesch. 33,370-412.

BEDDINGTON, R. S. P. (1981). An autoradiographic analysis of the potency of embryonicectoderm in the 8th day postimplantation mouse embryo. J. Embryol. exp. Morph. 64, 87-104.

BEDDINGTON, R. S. P. (1982). An autoradiographic analysis of potency in different regions of theembryonic ectoderm during gastrulation in the mouse. /. Embryol. exp. Morph. 69, 265-285.

BEDDINGTON, R. S. P. (1983a). Histogenetic and neoplastic potential of different regions of themouse embryonic egg-cylinder. J. Embryol. exp. Morph. 75,189-204.

BEDDINGTON, R. (19836). The origin of the foetal tissues during gastrulation in the rodent. InDevelopment in Mammals, vol. 5 (ed. M. H. Johnson), pp. 1-32. Amsterdam, New York,Oxford: Elsevier Science Publishers B.V.

BIJTEL, J. H. (1936). Die Mesodermbildungspotenzen der hinteren Medullarplattenbezirke beiAmblystoma mexicanum in Bezug auf die Schwanzbildung. Wilhelm Roux Arch. EntwMech.Org. 134, 262-282.

BODE, V. C. & DZIADEK, M. A. (1979). Plasminogen activator secretion during mouseembryogenesis. Devi Biol. 73, 272-289.

BRUNNER, G. & TSCHANK, G. (1982). Contracting striated muscle fibres differentiated fromprimary rat pituitary cultures. Cell Tissue Res. 224, 655-662.

BULGER, R. E. (1973). Rat renal capsule: presence of layers of unique squamous cells. Anat. Rec.177, 393-408.

BUTCHER, E. O. (1929). The development of the somites in the white rat (Mas norvegicus albinus)and the fate of the myotomes, neural tube, and gut in the tail. Am. J. Anat. 44, 381-439.

CLEINE, J. H. & SLACK, J. M. W. (1985). Normal fates and states of specification of differentregions in the axolotl gastrula. /. Embryol. exp. Morph. 86, 247-269.

22 A . SVAJGER, B . LEVAK-SVAJGER AND N. SKREB

CRILEY, B. B. (1969). Analysis of the embryonic sources and mechanisms of development ofposterior levels of chick neural tubes. /. Morph. 128, 465-502.

CURRIE, J. R., MAYLIE-PFENNINGER, M.-F. & PFENNINGER, K. H. (1984). Developmental^regulated plasmalemmal glycoconjugates of the surface and neural ectoderm. Devi Biol. 106,109-120.

D'AMICO-MARTEL, A. & NODEN, D. M. (1983). Contributions of placodal and neural crest cells toavian cranial peripheral ganglia. Am. J. Anat. 166, 445-468.

DAMJANOV, I. (1978). Development of teratomas from embryos transplanted into outbred adulthamsters. /. natn. Cancer Inst. 61, 911-915.

DAMJANOV, I. & SOLTER, D. (1974). Experimental teratoma. Current Topics Path. 59, 69-130.DE BEER, G. R. (1947). The differentiation of neural crest cells into visceral cartilages and

odontoblasts in Amblystoma, and re-examination of the germ-layer theory. Proc. Roy. Soc. B134,377-398.

DE BEER, G. (1958). Embryos and Ancestors. 3rd edition. Oxford: Oxford University Press.DETTLAFF, T. A. (1983). A study of the properties, morphogenetic potencies and prospective fate

of outer and inner layers of ectodermal and chordamesodermal regions during gastrulation, invarious Anuran amphibians. /. Embryol. exp. Morph. 75, 67-86.

DIWAN, S. B. & STEVENS, L. C. (1976). Development of teratomas from the ectoderm of mouseegg cylinders. J. natn. Cancer Inst. 57, 937-939.

DRENCKHAHN, D., UNSICKER, K., GRIESSER, G.-H., SCHUMACHER, U. & GROSCHEL-STEWART, U.(1978). Different myosins in myoid and endodermal reticular epithelial cells of the thymus.An immunocytochemical study using specific antibodies against striated and smooth musclemyosin. Cell Tissue Res. 187, 97-103.

DZIADEK, M. A. (1981). Use of glycine as a non-enzymic procedure for separation of mouseembryonic tissues and dissociation of cells. Expl Cell Res. 133, 383-393.

DZIADEK, M. & ADAMSON, E. (1978). Localisation and synthesis of alphafetoprotein inpostimplantation mouse embryos. /. Embryol. exp. Morph. 43, 289-313.

FONTAINE, J. & LE DOUARIN, N. M. (1977). Analysis of endoderm formation in the avianblastoderm by the use of quail-chick chimaeras. /. Embryol. exp. Morph. 41, 209-222.

FRANKE, W. W., GRUND, CH. , KUHN, C., JACKSON, B. W. & ILLMENSEE, K. (1982a). Formation ofcytoskeletal elements during mouse embryogenesis. III. Primary mesenchymal cells and thefirst appearance of vimentin filaments. Differentiation 23, 43-59.

FRANKE, W. W., SCHMID, E., SCHILLER, D. L., WINTER, S., JARASCH, E. D., MOLL, R., DENK, H.,JACKSON, B. W. & ILLMENSEE, K. (19826). Differentiation-related patterns of expression ofproteins of intermediate-size filaments in tissue and cultured cells. Cold Spring Harb. Symp.quant. Biol. 46, 431-453.

FUJIMOTO, H. & YAMAGISAWA, K. O. (1979). Effects of the T-mutation on histogenesis of themouse embryo under the testis capsule. /. Embryol. exp. Morph. 50, 21-30.

GARDNER, R. L. & ROSSANT, J. (1976). Determination during embryogenesis. In Embryogenesis inMammals. Ciba Found. Symp. 40 (new series), pp. 5-25. Amsterdam, Oxford, New York:Elsevier, Excerpta Medica, North Holland.

GARDNER, R. L. & ROSSANT, J. (1979). Investigation of the fate of 4-5 day post-coitum mouseinner cell mass cells by blastocyst injection. /. Embryol. exp. Morph. 52, 141-152.

GAUDECKER, B. VON & MULLER-HERMELINK, H. K. (1980). Ontogeny and organization of thestationary non-lymphoid cells in the human thymus. Cell Tissue Res. 207, 287-306.

GREENBURG, G. & HAY, E. D. (1982). Epithelia suspended in collagen gels can lose polarity andexpress characteristics of migrating mesenchymal cells. /. Cell Biol. 95, 333-339.

GROBSTEIN, C. (1952). Intra-ocular growth and differentiation of clusters of mouse embryonicshields cultured with and without primitive endoderm and in the presence of possibleinductors. /. exp. Zool. 119, 355-380.

HAMILTON, W. J., BOYD, J. D. & MOSSMAN, H. W. (1976). Human Embryology, 4th edn. London:The Macmillan Press Ltd.

HARRISON, R. G. (1933). Some difficulties of the determination problem. Am. Nat. 67, 306-321.HAY, E. D. (1968). Organization and fine structure of epithelium and mesenchyme in the

developing chick embryo. In Epithelial-Mesenchymal Interactions (ed. R. Fleischmajer &R. E. Billingham), pp. 31-55. Baltimore: Williams & Wilkins.

HAY, E. D. (1973). Origin and role of collagen in the embryo. Am. Zool. 13,1085-1107.


HAY, E. D. (1978). Fine structure of embryonic matrices and their relation to the cell surface inruthenium-red fixed tissues. Growth 42, 399-423.

HEASMAN, WYLIE, C. C., HAUSEN, P. & SMITH, J. C. (1984). Fates and states of determination ofsingle vegetal pole blastomeres of X. laevis. Cell 37,185-194.

HEATH, J. K. (1978). Mammalian primordial germ cells. In Development in Mammals, vol. 3 (ed.M. H. Johnson), pp. 267-298. Amsterdam, New York, Oxford: North Holland PublishingCompany.

HOGAN, B. L. M. & TILLY, R. (1981). Cell interaction and endoderm differentiation in culturedmouse embryos. /. Embryol. exp. Morph. 62, 379-394.

HOLMDAHL, D. E. (1935). Primitivstreifen beziehungsweise Rumpfschwanzknospe in Verhaltniszur Korperentwicklung. Z. mikrosk.-anat. Forsch. 38, 409-440.

HOLTFRETER, J. (1938). Differenzierungspotenzen isolierter Teile der Urodelengastrula. WilhelmRowcArch. EntwMech. Org. 138, 522-656.

INESTROSA, N. C. (1982). Differentiation of skeletal muscle cells in culture. Cell Struc. Func. 7,91-109.

ISHIZUYA-OKA, A. & MIZUNO, T. (1984). Intestinal cytodifferentiation in vitro of chick stomachendoderm induced by the duodenal mesenchyme. /. Embryol. exp. Morph. 82, 163-176.

JACKSON, B. W., GRUND, C , WINTER, S., FRANKE, W. W. & ILLMENSEE, K. (1981). Formationof cytoskeletal elements during mouse embryogenesis. II. Epithelial differentiation andintermediate-sized filaments in early post-implantation embryos. Differentiation 20, 203-216.

JACOBSON, A. G. & TAM, P. P. L. (1982). Cephalic neurulation in the mouse embryo analysedby SEM and morphometry. Anat. Rec. 203, 375-396.

JOLLY, J. & F£RESTER-TADIE, M. (1936). Recherches sur l'oeuf du rat et de la souris. ArchsAnat.microsc. Morph. exp. 32, 323-390.

JURAND, A. (1974). Some aspects of the development of the notochord in mouse embryos./. Embryol. exp. Morph. 32, 1-33.

KELLER, R. E. (1980). The cellular basis of epiboly: a SEM study of deep-cell rearrangementduring gastrulation in Xenopus laevis. J. Embryol. exp. Morph. 60, 201-234.

KETELSEN, U.-P. & WEKERLE, H. (1976). Thymus-derived muscle clones. An ultrastructuralanalysis of cell differentiation. Differentiation 5, 185-187.

KOBAYASHI, K. (1978). Fine structure of the mammalian renal capsule: the atypical smoothmuscle cell and its functional meaning. Cell Tissue Res. 195, 381-394.

KOLLAR, E. J. (1983). Epithelial-mesenchymal interactions in the mammalian integument: toothdevelopment as a model for instructive induction. In Epithelial-Mesenchymal Interactions inDevelopment (ed. R. H. Sawyer & J. F. Fallon), pp. 27-49. New York: Preger.

LANDSTROM, U. & L0VTRUP, S. (1978). Fate maps and cell differentiation in the amphibianembryo. An experimental study. /. Embryol. exp. Morph. 54, 113-130.

LAWSON, K. A. (1974). Mesenchyme specificity in rodent salivary gland development: theresponse of salivary epithelium to lung mesenchyme in vitro. /. Embryol. exp. Morph. 32,469-493.

LAWSON, K. A., MENESES, J. J. & PEDERSEN, R. A. (1983). Fate mapping in presomite mouseembryos by intracellular microinjection of horseradish peroxidase. /. Cell Biol. 97, 37a.

LE LIEVRE, C. S. & LE DOUARIN, N. N. (1975). Mesenchymal derivatives of neural crest -Analysis of chiameric quail and chick embryos. J. Embryol. exp. Morph. 34,125-154.

LENNON, V. A., PETERSON, S. & SCHUBERT, D. (1979). Neuroectoderm markers retained inphenotypical skeletal muscle arising from a glial cell line. Nature, Lond. 281, 586-588.

LEVAK-SVAJGER, B. & SKREB, N. (1965). Intraocular differentiation of rat egg-cylinders.J. Embryol. exp. Morph. 13, 243-253.

LEVAK-SVAJGER, B. & SVAJGER, A. (1971). Differentiation of endodermal tissues in homografts ofprimitive ectoderm from two-layered rat embryonic shields. Experientia 27, 683-684.

LEVAK-SVAJGER, B. & SVAJGER, A. (1974). Investigation on the origin of the definitive endodermin the rat embryo. /. Embryol. exp. Morph. 32, 445-459.

LEVAK-SVAJGER, B. & SVAJGER, A. (1979). Course of development of isolated rat embryonicectoderm as renal homograft. Experientia 35, 258-259.

LEVAK-SVAJGER, B., SVAJGER, A. & SKREB, N. (1969). Separation of germ layers in presomite ratembryos. Experientia 25,1311-1312.


LIBBUS, B. L. & Hsu, Y. CH. (1980). Sequential development and tissue organization in wholemouse embryos cultured from blastocyst to early somite stage. Anat. Rec. 197, 317-329.

LIEB, R. J. & DE PAOLA, D. P. (1981). Ultrastructural alteration of trypsin- and pancreatin-separated embryonic rabbit palate epithelium and mesenchyme. J. dent. Res. 60, 164-170.

LOVELL-BADGE, R. H., EVANS, M. J. & BELLAIRS, R. (1985). Protein synthetic patterns in the earlychick embryo. /. Embryol. exp. Morph. 85, 65-80.

L0VTRUP, S. (1974). Epigenetics - A Treatise of Theoretical Biology. London: John Wiley.L0VTRUP, S. (1975). Fate maps and gastrulation in Amphibia - a critique of current views. Can.

J. Zool. 53, 473-479.L0VTRUP, S. (1983). Epigenetic mechanisms in the early amphibian embryo: cell differentiation

and morphogenetic elements. Biol. Rev. 58, 91-130.MCLAREN, A. (1983). Studies on mouse germ cells inside and outside the gonad. /. exp. Zool.

228, 167-171.MADEN, M. (1977). The role of Schwann cells in paradoxical regeneration in the axolotl.

/. Embryol. exp. Morph. 41, 1-13.MASLOW, D. E. (1976). In vitro analysis of surface specificity in embryonic cells. In The Cell

Surface in Animal Embryogenesis and Development (ed. G. Poste & G. L. Nicholson),pp. 697-745. Amsterdam, New York and Oxford: North Holland Publishing Company.

MEIER, S. & JACOBSON, A. G. (1982). Experimental studies of the origin and expression ofmetameric pattern in the chick embryo. /. exp. Zool. 219, 217-232.

MEIER, S. & TAM, P. P. L. (1982). Metameric pattern of development in the embryonic axis ofthe mouse. I. The differentiation of the cranial segments. Differentiation 21, 95-108.

MITRANI, E. (1982). Primitive streak-forming cells of the chick invaginate through a basementmembrane. Wilhelm Roux'Arch, devl Biol. 191, 320-324.

NAKAMURA, H. & AYER-LE LIEVRE, CH. S. (1982). Mesectodermal capabilities of the trunk neuralcrest of birds. /. Embryol. exp. Morph. 70, 1-18.

NAKAO, T. & ISHIZAWA, A. (1984). Light- and electron-microscopic observations on the tail budof the larval lamprey (Lampetra japonica), with special reference to neural tube formation.Am. J. Anat. 170, 55-71.

NATHANSON, M. A. & HAY, E. D. (1980). Analysis of cartilage differentiation from skeletalmuscle grown on bone matrix. I. Ultrastructural aspects. Devi Biol. 78, 301-331.

NEW, D. A. T. (1966). Culture of Vertebrate Embryos. London: Logos Press.NICOLET, G. (1971). Avian gastrulation. Adv. Morphogen. 9, 231-262.NIEUWKOOP, P. D. (1973). The "organization center" of the amphibian embryo: its origin,

spatial organization, and morphogenetic action. Adv. Morphogen. 10, 1-39.O'HARE, M. J. (1978). Teratomas, neoplasia and differentiation: a biological overview. I. The

natural history of teratomas. Invest. Cell Pathol. 1, 39-63.OPPENHEIMER, J. M. (1940). The non-specificity of the germ layers. Q. Rev. Biol. 15, 1-27.OSMAN, M. & RUCH, J.-V. (1981). Reconstitution of the basement membrane from the inner and

outer dental epithelia of trypsin-isolated mouse molar enamel organs. J. Biol. Buccale 9,129-139.

PARANKO, J., PELLINIEMI, L. J. & FOIDART, J.-M. (1984). Epithelio-mesenchymal interface andfibronectin in the differentiation of the rat mesonephric and paramesonephric ducts.Differentiation 27, 196-204.

PETER, K. (1941). Die Genese des Entoderms bei den Wirbeltieren. Ergebn. Anat. EntwGesch.33, 231-249.

PETER, K. (1951). Die zweifache Entwicklung des Wirbeltierkorpers in finaler, erhaltungs-funktioneller Bedeutung. Z. mikrosk.-anat. Forsch. 57, 393-401.

PIERCE, G. B. (1966). The development of basement membranes of the mouse embryo. DeviBiol. 13, 231-249.

POELMANN, R. E. (1980a). The mouse embryo: the development from one to three germ layers.Ph.D. Thesis. Leiden.

POELMANN, R. E. (19806). Differential mitosis and degeneration patterns in relation to thealterations in the shape of the embryonic ectoderm of early post-implantation mouseembryos. /. Embryol. exp. Morph. 55, 33-51.

POELMANN, R. E. (1981a). The formation of the embryonic mesoderm in the early post-implantation mouse embryo. Anat. Embryol. 162, 29-40.


POELMANN, R. E. (19816). The head-process and the formation of the definitive endoderm in themouse embryo. Anat. Embryol. 162, 41-49.

POELMANN, R. E. & VERMEIJ-KEERS, CH. (1976). Cell degeneration in the mouse embryo: aprerequisite for normal development. In Progress in Differentiation Research (ed. N. Miiller-B6rat etal.), pp. 93-102. Amsterdam: North Holland Publishing Company.

PUCHTLER, H.,MELOAN,S.N.,BRAUCH,B. W. &GROPP, S. (1975). Myoepithelial cells in humanthymus: staining, polarization and fluorescence microscopic studies. Histochemistry 45,163-176.

ROSSANT, J. & PAPAIOANNOU, V. E. (1977). The biology of embryogenesis. In Concepts inMammalian Embryogenesis (ed. M. J. Sherman), pp. 1-36. Cambridge: MIT Press.

RUGH, R. (1954). The effect of ionizing radiations on amphibian development. /. cell comp.Physiol. 43 (Suppl. 1), 39-67.

RUSSELL, L. B. & RUSSELL, W. L. (1954). An analysis of the changing radiation response of thedeveloping mouse embryo. J. cell. comp. Physiol. 43 (Suppl. 1), 103-147.

SANDERS, E. J. (1984). Labelling of basement membrane constituents in the living chick embryoduring gastrulation. /. Embryol. exp. Morph. 79, 113-123.

SCHOENWOLF, G. (1984). Histological and ultrastructural studies of secondary neurulation inmouse embryos. Am. J. Anat. 169, 361-376.

SCHOENWOLF, G. C. & NICHOLS, D. H. (1984). Histological and ultrastructural studies on theorigin of the caudal neural crest cells in mouse embryos. /. comp. Neurol. 222, 496-505.

SKREB, N. (1961). Experimental analysis of maturation of the organization pattern in rat egg. InSymposium on the Germ Cells and Earliest Stages of Development (ed. S. Ranzi), pp. 291-303.Milano: Int. Embryol and Fondazione A. Basells.

SKREB, N. & BIJELI6, N. (1962). Effects of X-rays on the rat embryo during mesoderm formation.v Nature 193, 292-293.SKREB, N. & CRNEK, V. (1977). Tissue differentiation in ectopic grafts after cultivation of rat

embryonic shields in vitro. J. Embryol. exp. Morph. 42, 127-134.SKREB, N. & CRNEK, V. (1980). Development of embryo-derived teratomas in vitro. In

Differentiation and Neoplasia (ed. R. G. McKinnell, M. A. Di Bernandino, M. Blumenfeld &R. D. Bergad), pp. 283-289. Berlin, Heidelberg, New York: Springer Verlag.

SKREB, N. & FRANK, Z. (1963). Developmental abnormalities in the rat induced by heat shock.J. Embryol. exp. Morph. 11, 445-457.

SKREB, N. & LEVAK, B. (1960). Etude sur la differentiation des jeunes embryons du rat blancdans l'hetejogreffe endoculaire. C.r. Seanc. Soc. Biol. 154, 1750-1752.

SKREB, N. & SVAJGER, A. (1975). Experimental teratomas in rats. In Teratomas and Differentiation(ed. M. Sherman & D. Solter), pp. 83-97. New York: Academic Press.

SKREB, N., SVAJGER, A. & LEVAK-SVAJGER, B. (1971). Growth and differentiation of rat egg-cylinders under the kidney capsule. /. Embryol. exp. Morph. 25, 47-56.

SKREB, N., SVAJGER, A. & LEVAK-SVAJGER, B. (1976). Developmental potentialities of the germlayers in mammals. In Embryogenesis in Mammals. Ciba Found. Symp. 40 (new series),pp. 27-45. Amsterdam: Elsevier.

SMEDLEY, M. J. & STANISSTREET, M. (1984). Scanning electron microscopy of wound healing in ratembryos. /. Embryol. exp. Morph. 83, 109-117.

SNOW, M. H. L. (1978). Techniques for separating early embryonic tissues. In Methods inMammalian Reproduction (ed. J. C. Daniel, Jr), pp. 167-178. New York, San Francisco,London: Academic Press.

SNOW, M. H. L. (1981). Autonomous develpment of parts isolated from primitive-streak-stagemouse embryos. Is development clonal?/. Embryol. exp. Morph. 65 (Suppl.), 269-287.

SNOW, M. H. L. & TAM, P. P. L. (1979). Is compensatory growth a complicating factor in mouseteratology? Nature, Lond. 279, 555-557.

SOLTER, D., DAMJANOV, I. & KOPROWSKI, H. (1975). Embryo-derived teratoma: a model systemin developmental and tumor biology. In The Early Development of Mammals (ed. M. Balls &A. E. Wild), pp. 243-264. London: Cambridge University Press.

SOLTER, D., DAMJANOV, I. & SKREB, N. (1970). Ultrastructure of mouse egg-cylinders. Z. Anat.EntwGesch. 132, 291-298.v

SOLTER, D., DAMJANOV, I. & SKREB, N. (1973). Distribution of hydrolytic enzymes in early rat andmouse embryos - a reappraisal. Z. Anat. EntwGesch. 139, 119-126.


SOLURSH, M. & REVEL, J. P. (1978). A scanning electron microscope study of cell shape and cellappendages in the primitive streak region of the rat and chick embryo. Differentiation 11,185-190.

SPIEGELMAN, M. (1976). Electron microscopy of cell associations in T-locus mutants. InEmbryogenesis in Mammals. Ciba Found. Symp. 40 (new series), pp. 199-226. Amsterdam:Elsevier.

SPIEGELMAN, M. & BENNETT, D. (1974). Fine structural study of cell migration in the earlymesoderm of normal and mutant mouse (T-locus: t9/t9).J. Embryol. exp. Morph. 32,723-738.

STEVENS, L. C. (1967). The biology of teratomas. Adv. Morphogen. 6, 1-31.STEVENS, L. C. (1968). The development of teratomas from intratesticular grafts of tubal mouse

eggs. /. Embryol. exp. Morph. 20, 329-341.STEVENS, L. C. & HUMMEL, K. P. (1957). A description of spontaneous congenital testicular

teratomas in strain 129 mice. /. natn. Cancer Inst. 18, 719-747.SUGRUE, S. P. & HAY, E. D. (1981). Response of basal epithelial cell surface and cytoskeleton to

solubilized extracellular matrix molecules. /. Cell Biol. 91, 45-54.SUMIYA, M. (1976). Differentiation of the digestive tract epithelium of the chick embryo

cultured in vitro enveloped in a fragment of the vitelline membrane in the absence ofmesenchyme. Wilhelm Roux' Arch, devl Biol. 179, 1-17.

SVAJGER, A., KOSTOVIC-KNEZEVIC, LJ., BRADAMANTE, Z. & WRISCHER, M. (1985). Tail gutformation in the rat embryo. Wilhelm Roux' Arch, devl Biol. 194, 429-432.

SVAJGER, A. & LEVAK-SVAJGER, B. (1974). Regional developmental capacities of the ratembryonic endoderm at the head fold stage. J. Embryol. exp. Morph. 32, 461-467.

SVAJGER, A. & LEVAK-SVAJGER, B. (1975). Technique of separation of germ layers in ratembryonic shields. Wilhelm Roux Arch, devl Biol. 178, 303-308.

SVAJGER, A. & LEVAK-SVAJGER, B. (1976). Differentiation in renal homografts of isolated parts ofrat embryonic ectoderm. Experientia 32, 378-379.

SVAJGER, A., LEVAK-SVAJGER, B., KOSTOVIC-KNEZEVIC, LJ. & BRADAMANTE, Z. (1981).Morphogenetic behaviour of the rat embryonic ectoderm as a renal homograft. /. Embryol.exp. Morph. 65 (Suppl.), 243-267.

SVAJGER, A., LEVAK-SVAJGER, B., KOSTOVIC-KNEZEVIC, LJ. & SKREB, N. (1985). Developmentalcapacities of the primitive embryonic ectoderm of the rat in renal isografts. Ada biologica(Yugoslav Academy of Sciences and Arts) 11, 21-33.

TAM, P. P. L. (1981). The control of somitogenesis in mouse embryos. J. Embryol. exp. Morph.65 (Suppl.), 103-128.

TAM, P. P. L. (1984). The histogenetic capacity of tissues in the caudal end of the embryonic axisof the mouse. /. Embryol. exp. Morph. 82, 253-266.

TAM, P. P. L. & MEIER, S. (1982). The establishment of a somitomeric pattern in the mesodermof the gastrulation mouse embryo. Am. J. Anat. 164, 209-225.

TAM, P. P. L., MEIER, S. & JACOBSON, A. G. (1982). Differentiation of the metameric pattern inthe embryonic axis of the mouse. II. Somitomeric organization of the presomitic mesoderm.Differentiation 21, 109-122.

TAM, P. P. L. & SNOW, M. H. L. (1981). Proliferation and migration of primordial germ cellsduring compensatory growth in mouse embryos. /. Embryol. exp. Morph. 64, 133-147.

TAMARIN, A. & BOYDE, A. (1976). Three-dimensional anatomy of the 8-day mouse conceptus: astudy by scanning electron microscopy. /. Embryol. exp. Morph. 36, 575-596.

THIERY, J.-P., DELOUVEE, A., GALLIN, W. J., CUNNINGHAM, B. A. & EDELMAN, G. M. (1984).Ontogenetic expression of cell adhesion molecules: L-CAM is found in epithelia derived fromthe three primary germ layers. Devl Biol. 102, 61-78.

TORO, E. (1938). The homeogenetic induction of neural folds in rat embryos. J. exp. Zool. 79,213-236.

TOWNES, P. L. & HOLTFRETER, J. (1955). Directed movements and selective adhesion of embryonicamphibian cells. /. exp. Zool. 128, 53-120.

TRELSTAD, R. L., HAYASHI, A. A., HAYASHI, K. & DONAHOE, P. K. (1982). Theepithelial-mesenchymal interface of the male rat Mullerian duct: Loss of basementmembrane integrity and ductal regression. Devl Biol. 92, 27-40.


UNO, K. (1982). A correlation between the capacity of cavity formation and the subsequentdifferentiation of terato-carcinoma embryoid body lines. /. Embryol. exp. Morph. 69,127-140.

USADEL, K. H., ROCKERT, H., OBERT, I. & SCHOFFLING, K. (1970). Uber dieEntwicklungisologerZellsuspensions - Transplantate aus embryonalen Darm- und Magenanlagen bei der Ratte.Z. ges. exp. Med. 153, 273-289.

VEINI, M. & HARA, K. (1975). Changes in the differentiation tendencies of the hypoblast-freeHensen's node during "gastrulation" in the chick embryo. Wilhelm RouxArch. devlBiol. 177,89-100.

VERMEIJ-KEERS, CH. & POELMANN, R. E. (1980). The neural crest: a study on cell degenerationand the improbability of cell migration in mouse embryos. Neth. J. Zool. 30, 74-81.

VERWOERD, C. D. A. & OOSTROM, C. G. VAN (1979). Cephalic neural crest and placodes. Adv.Anat. Embryol. Cell Biol. 58, 1-75.

VOGT, W. (1929). Gestaltungs analyse am Amphibienkeim mit ortlicher Vitalfarbung. II. Teil.Gastrulation und Mesodermbildung bei Urodelen und Anuren. Wilhelm Roux Arch.EntwMech. Org. 120, 384-706.

WADDINGTON, C. H. (1936). Organizers in mammalian development. Nature, Lond. 138, 125.WADDINGTON, C. H. (1957). Principles of Embryology. London: Georg Allen & Unwin.WALLACE, H. (1972). The components of regrowing nerves which support the regeneration of

irradiated salamander limbs. /. Embryol. exp. Morph. 28, 419-435.WARTIOVAARA, J., LEIVO, I. & VAHERI, A. (1979). Expression of the cell associated glycoprotein,

fibronectin, in the early mouse embryo. DevlBiol. 69, 247-257.WAYMOUTH, C. (1974). To disaggregate or not to disaggregate. Injury and cell disaggregation,

transient or permanent. In vitro 10, 97-111.WEKERLE, H., PATERSON, B., KETELSEN, U.-P. & FELDMAN, M. (1975). Striated muscle fibers

differentiate in monolayer cultures of adult thymus reticulum. Nature, Lond. 256, 493-494.WILEY, L. M. & PEDERSEN, R. H. (1977). Morphology of mouse egg-cylinder development

in vitro: a light and electron microscopic study. /. exp. Zool. 200, 389-402.WILEY, L. M., SPINDLE, A. I. & PEDERSEN, R. H. (1978). Morphology of isolated mouse inner cell

mass developing in vitro. DevlBiol. 63,1-10.WILLIS, R. A. (1935). The structure of teratoma. /. Path. & Bad. 40, 1-36.WILLIS, R. A. (1962). The Borderland of Embryology and Pathology, pp. 422-463. London:

Butterworths.WILSON, J. G. (1954). Differentiation and the reaction of rat embryos to radiation. /. Cell Comp.

Physiol. 43 (Suppl. 1), 11-37.YAMADA, T. (1977). Control mechanisms in cell-type conversion in newt lens regeneration.

Monographs in Developmental Biology, vol. 13. Basel, New York: Karger.YASUGI, S. & MIZUNO, T. (1978). Differentiation of the digestive tract epithelium under the

influence of the heterologous mesenchyme of the digestive tract in the bird embryos. DeviGrowth and Differ. 20, 261-267.

ZIMMERMANN, B., MERKER, H.-J. & BARRACH, H.-J. (1982). Basement membrane alterationsafter treatment with trypsin, hyaluronidase and collagenase. Virchows Arch. (Cell Pathol.) 40,9-15.

rat embryonic ectoderm as renal isograft

Documents