The chicken blastoderm: current views on cell biological events guiding intercellular communication

Cell Differentiation, 22 (1988) 83-106 83 Elsevier Scientific Publishers Ireland, Ltd. CDF 00474 Review The chicken blastoderm: current views on cell biological events guiding intercellular communication Fernand Harrisson ~, Luc Andries ~ and Lucien Vakaet 2 I Department of Anatomy and Embryology, State University of Antwerp. Belgium, and : Department of Anatomy and Embryology, State University of Ghent, Belgium (Received 23 October 1987) Chicken blastoderm; Gastrulation; Intercellular communication; Cell junctions; Extracellular matrix; Cytoskeleton; Cell surface; Electrical fields Introduction It is generally believed that synchronized development is determined by the existence of an internal clock mechanism that regulates the timing of the morphogenetic events (Goodwin and Cohen, 1969) and of chemical messages that provide positional information to the cells (Wolpert, 1977). Timing of morphogenetic events such as cell division, determination of cleavage planes and of polarity appears to be primarily an intracellular phenomenon, whereas transmission of information probably requires different forms of intercellular communication and proceeds via message molecules. According to Saxrn (1980), intercellular communication consists of both short-range (requiring close apposition of cells) and long-range (over distances of the order of 50 000 nm) interactions between ceils. At the level of morphogenetic fields, i.e. groups of cells that possess a common developmental fate, short-range cell communication implies direct physical contact between cells and may occur by the interaction of cell-surface Correspondence address: Dr. F. Harrisson, State University Centre of Antwerp, Department of Anatomy and Embryology, 171 Groenenborgerlaan, B-2020 Antwerpen, Belgium. molecules (Moscona, 1974; Kemler et al., 1980), by short-range diffusion of signals (Weiss and Nir, 1979) or by transmission of molecules directly from cell to cell (Saxrn, 1977). In the last case, the transfer of information occurs through distinct intercellular membrane specializations (Revel, 1974). The widespread existence of low-electricalresistance junctions between embryonic cells may argue in favour of this mechanism, and their role in short-range intercellular communication has been reviewed (Douglas Powers and Tupper, 1977). Long-range intercellular communication between different morphogenetic fields includes free diffusion of molecules through the external milieu of the embryo, and the involvement of extracellularmatrix molecules other than the one belonging to the cell surface proper. The experimental disruption of these interactions leads to impaired morphogenesis. Among the different developmental systems that have been investigated, the chicken embryo is probably one of the best documented. It is the purpose of this review to bring together relevant literature and some unpublished observations on cell biological processes guiding intercellular communication in early chicken development. Our understanding of the processes which lead cells to 0045-6039/88/$03.50 © 1988 Elsevier Scientific Publishers Ireland, Ltd. 84 interact represents the synthesis of many lines of investigation. Apart from the study of cell junctions, which obviously mediate short-range transmission of molecules, it may also seem appropriate to collect more information about the extracellular matrix, the cytoskeleton and its relation to the extracellular space, the cell surface and its microappendages, as well as electric fields between tissues. For a better insight into these basic biological phenomena, the morphology of the chicken blastoderm has been overviewed briefly. Unpublished topographical schemes of the different developmental stages have been included in this report, to delineate clearly the different anlage fields, to introduce the terminology used, and to provide some aid to unexperienced researchers. Fate maps have been published earlier by one of us (Vakaet, 1984b, 1985). Morphology of the chicken blastoderm After laying, the incubation of fertilized eggs initiates the process of gastrulation in the blastoderm. The initial stages of development thus proceed before laying. These early stages of development, which include the period of cleavage, have been described by Eyal-Giladi and Kochav (1976). At the top of the yolk of undisturbed, fertilized eggs floats the blastoderm that, at the time of laying, appears as a disc of about 3 mm in diameter. The blastodisc is composed of two regions that are recognizable in whole mounts examined by transmitted light: a thin, central region, the area pellucida, and a thicker peripheral region, the area opaca. The former region gives rise to the embryo, and the latter to the yolk sac. The yolk and the blastoderm are enclosed within a thin, two-layered vitelline membrane. Differentiation of the area opaca The area opaca or germ wall is composed of an upper layer, the extraembryonic epiblast, which is adjacent to the vitelline membrane and continuous with the upper layer of the area pellucida. Beneath this layer, the yolk endoderm contains more and larger yolk droplets, which confer a opaque aspect to this region of the blastodisc. The yolk endoderm is composed of loosely attached cells moving laterally along the basement membrane of the extraembryonic epiblast. From about the tenth hour of incubation and during the first days of incubation, the extraembryonic epiblast migrates by active locomotion of a band of cells at the edge of the epiblast, where it is attached to the inner layer of the vitelline membrane (New, 1959). This band of cells has been referred to as the edge cells (Downie and Pegrun, 1971) or margin of overgrowth (Bellairs and New, 1962). The margin of overgrowth, which has been regarded as a study model for cell-to-cell contacts (Mareel and De Ridder, 1970), gradually grows around the yolk, to enclose it completely after 4 days of incubation. This phenomenon of epiboly is responsible for the radial tension that is exerted on the area pellucida, and that is necessary for normal development to occur (Kucera and Burnand, 1987). Descriptive and experimental work has clarified the structure and function of the margin of overgrowth (Andries et al., 1983a,b, 1985a; Andries and Vakaet, 1985a,b). Differentiation of the area pellucida The development of the area pellucida during the first 24 h of incubation is divided, according to Vakaet (1970), into 10 stages, from 0 through 9, recognizable on the basis of changes in shape. These changes are attributable to changes in the morphology of the different germ layers. The terminology used includes the terms 'upper layer' (or epiblast), 'middle layer' (or mesoblast), and 'deep layer' (composed of endophyll, hypoblast, and definitive endoblast), to refer to the relative position of the germ layers within the embryo without presuming the fate of the cells of a given layer. The major morphogenetic movements that occur within the area pellucida during gastrulation are the formation of the primitive streak and the establishment of bilateral symmetry (Eyal-Giladi, 1984, and Bellairs, 1986, for review), the formation of mesoblast (Sanders, 1986a, for review), and the formation of hypoblast and of definitive endoblast (Bellairs, 1982 and Eyal-Giladi, 1984, for review). Gastrulation in birds is defined as the process by which cells from the upper layer move 85 i, , +i! various lengths of time. The time necessary to obtain a given stage is, however, largely dependent on several external factors such as the temperature of incubation, the strain of chicken used, the period of the year, and the delay between laying and beginning of the incubation. Mean values applicable to the chicken are given in Fig. 2, which also compares our stages with those introduced by Hamburger and Hamilton (1951) and by EyalGiladi and Kochav (1976). A topographical scheme summarizing the structure of the blastoderm, as well as a sagittal section and transversal sections, is added for each stage (Fig. 3), i~ + 1 A series of stages in chicken gastrulation Stage O. This stage is found in freshly laid, unincubated eggs of about 20-h uterine age. The differentiation of the blastodisc into area opaca and area pellucida is completed at laying. The cells of the upper layer are arranged to form an epithelial sheet, whereas the deep layer or endophyll is composed of clusters of cells adhering to the basal surface of the upper layer. Stage 1. Changes of shape in the area pellucida are not observed during the first hours of incubation of the fertilized egg. The first reaction to incubation temperature is a centripetal movement Fig. 1. Schematic representation of the morphogenetic movements occurring in the area pellucida of a gastrulating stage-6 embryo. The arrows indicate the sense of insertion of definitive endoblast cells in the deep layer (DL), prior to stages 5-6. and the sense of migration of mesoblast cells (ML). to the interior of the germ, by ingression along an anteroposterior axis referred to as the primitive streak, forming definitive endoblast and mesoblast (see also Trinkaus, 1976). The morphogenetic movements are shown in Fig. 1. The stages are attained by incubating the freshly laid egg for GASTRULATION 1 I ". "'XI! ~o XI-XIV 20 hours of uterine oge 1 1 2 I 2 I 3 I 3 J 4 5 l t 4 l 6 I 5 I 7 I 8 I 6 l H~burger ond Homllton 1951 9 I Vokoet 1970 I EyoI-Gllodl ond Kochov 1976 ~D ! 12 I 18 1 24 hours of |ncuhotion 2 Fig. 2. Comparison of stages distinguished by different authors during the first 24 h of incubation. The scheme compares the different stages on the basis of the theoretical number of hours, reported by the authors, to reach the stages. However, in practice, the stages 3 and 4 of Hamburger and Hamilton are, respectively, very similar to the stages 3 and 7 of Vakaet. The stages 4-6 have been introduced by the latter author, mainly to distinguish major events occurring during this important period of development. 86 1_. ,.. ~ \ StageO i f _ . . . . . . . . i ~Z______~ 2 i ~° i ~'. 4 3 Fig. 3. Schematic representation of a series of stages in chicken gastrulation. AO, area opaca; AP, area pellucida; DL, deep layer composed of yolk endoderm (1), endophyll (2), hypoblast (3), and definitive endoblast (4): HF, head fold; HP, head process; MI. middle layer: MOv, margin of overgrowth; PS, primitive streak: UL, upper layer. 87 -.ql ~\ i i / ....-=-~i 0S ~ ~ " ..... -I i .................... ! 6 -.41 2 ! • "L a a 3 4 : , i Ii ii : I Fig. 3 (continued). 88 originating from the inner margin of the yolk endoderm. This slight convergence of deep-layer cells is accompanied by a condensation in the area pellucida, the area centralis. The junctional area which encircles it is thinner and more transparent when observed with transmitted light. The area centralis appears denser due to a thickening of the upper layer, but a concentration of endophyll material certainly contributes to its appearance. The border separating junctional and central areas of the area pellucida is poorly defined initially. It becomes clearly demarcated at the posterior end of the central area, due to the appearance of a sickle-shaped thickening, the sickle of Koller (1882). At the level of Koller's sickle develops the hypoblast, which gradually penetrates the central area and replaces the endophyll in a posterior-toanterior direction. Polyingression of single cells from the upper layer seems to represent a major morphogenetic movement in the formation of the hypoblast (Kochav et al., 1980; Eyal-Giladi, 1984; Weinberger et al., 1984). Stage 2. This stage is characterized by the appearance of the primitive streak which crosses the middle of the sickle. The anterior end moves anteriorly, whereas the posterior end moves posteriorly. Upper-layer cells de-epithelize at this level and form the middle layer. The hypoblast, situated beneath the single middle-layer cells, further extends along a posteroanterior axis. The endophyll is now observed as a continuous sheet, separated from the upper layer. Stage 3. The elongation of the primitive streak, which is now very rapid, brings about a change in the shape of the area pellucida: the posterior end of the streak extends into the area opaca, over the yolk endoderm, conferring a pear-shaped aspect to the area pellucida. The anterior part of the primitive streak does not yet reach the centre of the central area, which is regarded as a punctumfixum. A longitudinal groove at the level of the primitive streak is not yet present at the apical surface of the upper layer. The middle layer extends laterally and the hypoblast further moves anteriorly. Stage 4. During this stage, the primitive streak elongates in one direction, towards the centre of the central area, and in the opposite direction, over the yolk endoderm, and it becomes thicker. The pear shape of the area pellucida is more marked. A groove appears at the level of the primitive streak at the end of this developmental period. In the deep layer, a third generation of cells appears: the definitive endoblast. The cells of this layer originate from the upper layer, by ingression at the anterior end of the primitive streak, and insert in the hypoblast (Vakaet, 1962). Stage 5. This stage is characterized by a clear-cut groove in the primitive streak, which initially appears in the anterior half and gradually extends posteriorly. From now on, the anterior end of the primitive streak is referred to as Hensen's node. The posterior end is still hidden by yolk endoderm. The ingressed upper-layer cells that will give rise to the definitive endoblast, move away from the node in a concentric fashion, pushing the hypoblast more laterally. The posterior part of the area pellucida, as well as part of the area opaca, is colonized by the middle layer. Stage 6. During this stage, the length of the primitive streak is maximal. Hensen's node has reached its most anterior position, and the posterior end of the primitive streak is still situated in the area opaca. Around the node, an almost concentric condensation in the upper layer corresponds to the neurectoblast. Expansion of definitive endoblast and of mesoblast becomes more and more evident. Stage 7. From this stage on, a shortening of the length of the primitive streak occurs. Hensen's node regresses posteriorly and the posterior end of the primitive streak is now often visible in the area pellucida. A primordium of the head process appears as the anterior half of the primitive streak regresses. Stage 8. This stage is marked by the presence of a head process surrounded rostrally and laterally by the neurectoblast condensation. Shortening of the primitive streak is evident. Stage 9. This stage is easily identified by the appearance of the head fold in front of the rostral extremity of the head process. Due to the expansion of the hypoblast and of the definitive endoblast, the endophyll is now restricted to a crescent situated in the anterior part of the area pellucida. These events precede the processes of neurulation and somitogenesis, which are characterized by the 89 elevation and closure of the neural folds, and segregation of the first pair of somites. Patterns of junctional communication Some aspects of short-range intercellular communication involve the presence of cell junctions. In unincubated chicken blastoderms (stage 0), which consist of a complete epithelial upper layer (epiblast) and an incomplete deep layer (endophyll), focal tight junctions and occasionally tight junctions are the earliest junctions observed (Trelstad et al., 1967; Sanders, 1973; Bellairs et al., 1975). They are mainly found at the apical pole of the epiblast cells. Desmosomes and micropapillae have been encountered in deeper contact regions between these cells (Sanders, 1973), although it has been reported earlier that desmosomes were not discovered in primary epithelia (Balinsky and Walther, 1961). Gap junctions are extremely uncommon (Bellairs et al., 1975). Between deep-layer cells, focal tight junctions (maculae occludentes) (Trelstad et al., 1967) and micropapillae (Sanders, 1973) have been described. Using reaggregates of unincubated chicken embryos, Macarak (1976) reported that, in 48-h reaggregates, the epiblast cells are joined by unspecialized junctions, while deep-layer cells are joined by desmosomes. The formation of desmosomes has been correlated with the sorting out process of the dissociated ceils. This junctional pattern does not necessarily reflect the in vivo situation. During the first hours of incubation, gap junctions are more frequently encountered in the upper layer (Bellairs et al., 1975). The primary mesoblast cells that migrate laterally away from the primitive streak, make close and focal tight junctions with each other and with the basal surfaces of upper and deep layers (Trelstad et al., 1966; 1967; Revel, 1974; Revel and Solursh, 1986). Their occurrence in migrating tissues is, according to Trelstad et al. (1967), particularly interesting because of the possibility that tight junctions, acting as pathways of low-electrical resistance between cells, may mediate migration of mesoblast cells. The number of such focal tight junctions is low in migrating cells and increased after microinjection of hyaluronidases, which provoke compaction of cells, suggesting that detachment and migration of mesoblast cells is mediated by the presence of hyaluronate (Van Hoof et al., 1986). During late gastrulation, well-developed cell junctions including gap junctions (Bellairs et al., 1975) and, from apical to basal, tight junctions (zonulae occludentes), zonulae adhaerentes, and desmosomes (maculae adhaerentes) have been found (Figs. 4 and 5) (Overton, 1962, 1975, for review). Focal tight junctions and gap junctions, but not desmosomes and zonulae adhaerentes, are present between mesoblast cells (Hay, 1968; Revel et al., 1973). In embryos grown for 24 h in culture, the hypoblast displays tight junctions only, whereas the definitive endoblast contains gap junctions and desmosomes as well (Stagno and Low, 1980; Stolinsky et al., 1981). These differences have been related to the cohesiveness of the deep layer, which is low in hypoblast and high in definitive endoblast (Vakaet and Hertoghs-De Maere, 1973; Sanders et al., 1978). The description of cell junctions at the level of the margin of overgrowth, and hence their role in intercellular communication, has been greatly advanced in recent years through the use of freezefracture and electron microscopy of replicas. The blastoderm edge (New, 1959) or margin of overgrowth (Bellairs, 1963) migrates by locomotion of epithelial edge cells of the area opaca on the vitelline membrane used as a substrate. Assuming that the rapid migration of the marginal cells must be accompanied by coordination of the movement of the individual cells, Andries et al. (1985a) thoroughly investigated the spatial and temporal distribution of cell junctions in the margin of overgrowth. They found that tight junctions, gap junctions, and desmosomes are distributed in a pattern that gives information about the mechanism of locomotion of cells. Taking into account the distribution as well as the duration of formation and disassembly of cell junctions, they concluded that only the cells adhering to the vitelline membrane have the capability to change their position. The pattern of cell junctions reported by these authors makes less probable the movement of the ventral cells of the margin of overgrowth over the surface 90 Fig. 4. Transmission electron micrograph of the apical surface of the upper layer of the area pellucida. The blastoderm has been fixed in a hypotonic buffer solution containing glutaraldehyde and tannic acid, to increase the visualization of intercellular junctions and of microfilament bundles (arrowhead). The arrows indicate the presence, from apical to basal, of zonulae adhaerentes and of a gap junction ( × 37 570). Fig. 5. Freeze-fracture micrograph of the apical surface of the upper layer of the area pcllucida of a stage-6 blastoderm. This picture shows the P-face of tight junctions (TJ), desmosomes (D). and a small gap junction (G J) ( × 83 110). 91 of the cells adhering to the vitelline membrane, i.e. the roiling and sliding model of epithelial migration. Andries and Vakaet (1985b) clearly showed that gap junctions are present in the margin of overgrowth during gastrulation, but are virtually absent in the non-marginal epiblast. The fact that gap junctions have been implicated in intercellular communication, metabolic coupling, and cellto-cell transfer of metabolites suggests that the presence of these junctions is one of the factors that allow cooperation between marginal cells, so that a morphological unit is maintained and the movement of the individual cells synchronized. Extracellular matrix A large number of studies has documented the importance of the extracellular matrix in morphogenesis of several systems in general (Slavkin, 1986, for references) and of the chicken embryo in particular (Sanders, 1986b, for review). Immunocytochemical and experimental analyses of extracellular matrices have, indeed, demonstrated their implication in a variety of cell processes including cell-to-cell and cell-to-substrate adhesion, specific binding to biological macromolecules and to receptors of the cell surface, promotion of cell migration and positioning, transmembrane triggering of intracellular events, and modulation of cell shape. Because of the importance of such processes in development, a number of laboratories have focussed their research programme on this aspect of intercellular communication. The precise localization and the determination of the nature of the molecules have shown that most of the extracellular components found in adult tissues are already present in the embryo, some of them being expressed during gastrulation. At that time, four main classes of macromolecules are found in the matrix. These are (1) glycosaminoglycans (Abrahamsohn et al., 1975; Manasek, 1975; Solursh, 1976; Fisher and Solursh, 1977; Sanders, 1979; Wakely and England, 1979; Vanroelen et al., 1980a,b,c; Vanroelen and Vakaet, 1981), represented in major proportion by hyaluronate, but also by chondroitin sulphates and heparan sulphate; (2) non-collagenous glycopro- teins, including fibronectin (Critchley et al., 1979; Wakely and England, 1979; Duband and Thiery, 1982; Mitrani, 1982; Mitrani and Farberov, 1982; Sanders, 1982; Harrisson et al., 1984b, 1985c; Monnet-Tschudi et al., 1985) and laminin (Mitrani and Farberov, 1982), and their receptors (Krotoski et al., 1986); (3) types I and III interstitial collagens and basement membrane type IV collagen (Manasek, 1975); and (4) cell adhesion molecules (Thiery et al., 1982, 1984; Edelman et al., 1983). These components have been observed as amorphous material, surrounding mainly the mesenchymal cells and to a lesser extent the deeplayer cells, and as structurally well-defined material assembled to form a basement membrane beneath the epiblast. In spite of a large variability in the chemical composition of the basement membrane, it is generally believed that the cells that are underlaid by a basement membrane are responsible for the biosynthesis and the assembly of its components. In recent years, however, the use of chicken/quail chimaeras (Le Douarin, 1974) has greatly increased our knowledge of the role of underlying tissue in the formation of the basement membrane of the epiblast. The transplantation of metabolically labelled quail deep layers into chicken blastoderms and subsequent autoradiographic analysis of the chimaeras have demonstrated the participation of the deep layer in the synthesis of, at least, basement-membrane glycoproteins (Harrisson, 1986; Harrisson et al., 1985b). These studies, which have shown a dual cellular origin of basement-membrane components, lay stress on the importance of cell-cell interactions in development. Attempts to correlate morphogenetic events with the expression and distribution of extracellular material components have been made. For example, the formation of mesoblast involves a complex sequence of events that may be described successively as the formation of the primitive streak (Bellairs, 1986, for review), the de-epithelization of upper-layer cells at the level of the primitive streak (Vakaet et al.. 1980: Vakaet, 1984a) and their ingression through the basement membrane (Mitrani, 1982), and the lateral migration of mesoblast cells (Sanders, 1986a, for review). C t~ " ~ r ~~ *I F. ~ 8P D r~ "L ~ CD 0" 93 Especially the last feature largely depends on the available substrate (Fisher and Solursh, 1979a,b; Sanders, 1980), and it has frequently been suggested that mesoblast cells use the basement membrane as a natural substrate for adhesion and migration (Jacob et al., 1974; Revel, 1974; Ebendal, 1976; England, 1981; Sanders, 1983, for review). From various studies, it has become evident that the migration of mesoblast cells is mediated, at least, by the presence of hyaluronate and of fibronectin. The spatial and temporal organization of hyaluronate favours this idea. The phase of de-epithelization of upper-layer cells occurs at a site in which synthesis of hyaluronate is high (Fig. 6) (Vanroelen et al., 1980b). The cells are densely packed, intercellular junctions occur, and cell processes are few in number and in size (Van Hoof et al., 1986). During the lateral migration, increasing amounts of hyaluronate are present. This increase is reflected in the distribution of alcian-blue staining in primitive-streak-stage blastoderms, i.e. the periphery of middle-layer cells is negative at the level of the primitive streak, but increasingly positive in the lateral region (Fig. 7) (Vanroelen et al., 1980c). The cells of the latter region possess few intercellular junctions, and they extend long, thin cell processes. It is believed that hyaluronate, by its physicochemical properties, plays an essential role in creating spaces (Toole, 1982). In the chicken blastoderm, this is confirmed experimentally by Fisher and Solursh (1977) and by Van Hoof et al. (1984, 1986), who observed a drastic decrease in the cell-free spaces between mesoblast cells, respectively after a sub-blastodisc injection and after microinjection of hyaluronidases (Fig. 8). Ultrastructural investigation demonstrated that the removal of hyaluronate, not only affects the cell shape and leads to compaction of the ceils, but also induces the retraction of cell processes and the formation of intercellular junctions and, consequently, the loss of the mesenchymal aspect of the middle layer. This effect is reversed after a few hours of incubation, and normal development further proceeds. Fibronectin has been implicated for many years in the adhesion and migration of cells on substrates (Thiery, 1985, for review). The presence of fibronectin in the basement membrane of the chicken epiblast has been demonstrated ultrastructurally (Figs. 9 and 10) (Sanders, 1982; Harrisson et al., 1985a; Monnet-Tschudi et al., 1985), but experimental evidence for the implication of this glycoprotein in the migration of mesoblast cells is lacking until now. In a series of experiments, we performed microinjection of antifibronectin antibodies into blastoderms of stages 4-6 (to be published). Light- and electron microscope observation of these blastoderms after culture of several hours have revealed that the de-epithelization is not interrupted by the treatment, but that the lateral migration is inhibited in areas of rapid ingression. This leads to the formation of blastoderms in which the anterior part of the primitive streak is very thick, due to the accumulation of ingressed cells that do not seem to encounter an adequate substrate for lateral migration. Finally, it appears that hyaluronate, which is the major glycosaminoglycan in gastrulation (Fisher and Solursh, 1977; Vanroelen et al., 1980a), is complexed with fibronectin in areas of the basement membrane used by mesoblast cells for migration. Indeed, an interaction between hyaluronate and fibronectin has been demonstrated at the level of the epithelial-mesenchymal inter- Fig. 6. Autoradiograph of a stage-6 blastoderm after incorporation of tritiated glucosamine for 1 h. Positivity is localized at the periphery of ingressing middle-layer cells (a), and along the basement membrane (b). A lower grain density is observed in the vicinity of middle-layer cells starting their emigration from the primitive streak (PS) ( x 200). Fig. 7. Alcian-blue staining of a stage-6 blastoderm showing the accumulation of glycosaminoglycans during the migration ot mesoblast cells. The cells are negative at the level of the primitive streak (PS) (a), but increasingly positive in the lateral region (b) ( x 385). Fig. 8. Photomicrographs of semi-thin sections showing the typical mesenchymal aspect of the middle layer in the lateral region of the area pellucida (a), and the compaction of the middle layer after microinjection of testicular hyaluronidase (b) ( x 360). @ ." , .~.. ~. 'tl lID j' i I f ~;'~ • ,.{, ~',lJlD'~ilb,". "~''~-~ ,.'~,,~,L. ~.4 "b ~ , 95 face of the basement membrane (Harrisson et al., 1984a, 1985a), and this interaction is spatially and temporally correlated with the migration of mesoblast cells (Harrisson et ai., 1984b). This suggests that the mesenchyme is involved in some way in the interaction between both molecules and stresses the importance that mesenchymal tissues may have in the structural organization of a basement membrane. Direct experimental evidence in support of a remodelling of basementmembrane components mediated by migrating mesoblast cells has, moreover, been given (Van Hoof and Harrisson, 1986). In conclusion, the involvement of epithelialmesenchymal interactions and of cell-matrix interactions in the control of developmental processes, which has been demonstrated in several systems, is also evident in chicken gastrulation. Cytoskeletal elements Ingression of cells during chicken gastrulation is announced by a thickening of the cell layer that arises as upper-layer cells elongate along their apicobasal axis at the level of the primitive streak and modify their pyramidal shape to become bottle-shaped (Revel and Solursh, 1978; Solursh and Revel, 1978). Such changes of cell shape, together with loss of basement membrane as well as changes in adhesiveness between ingressing and non-ingressing cells, have been considered to be directly responsible for ingression (Bellairs, 1986, for review). Several views have been expressed concerning the causes of shape changes. Cytoskeletal elements, such as microfilaments and microtubules, may be responsible for changes in cellular form (Trinkaus, 1969, for review). However, taking into account the distribution of microfilaments in the upper layer, one has to be aware of the fact that these structures are ubiquitous in the upper layer. Indeed, 4-6-nm microfilament bundles have been localized near the apical surface of all upper-layer cells, in close association with cell junctions (Fig. 4), and these microfilaments are not restricted to the primitive-streak region (Vakaet and Vanroelen, 1982). In fact, a ring of actin filaments anchored to the apical cell junctions encircles the cells of the upper layer, forming a network that may provide stability and cohesion to this layer (Wakely and Badley, 1982). Such polygonal networks (Fig. 11), composed of actin microfilaments and associated with substrate-attached retraction processes, have also been described in explants and in dissociated cells of early chicken embryos (Ireland and Voon, 1981; Ireland et ai., 1983). Stimuli, which may cause contraction of microfilament bundles, must be considered to explain the formation of bottle-shaped cells in the primitive-streak region. Cyclic nucleotides are directly involved in the control of microfilament behaviour (Clarke and Spudich, 1977, for review). Evidence has been provided that the primitive streak elongates towards an artificial source of Fig. 9. lmmunocytochemical demonstration of fibronectin in the basal lamina, interstitial bodies, and along the basal surface of the upper layer ( × 34580). Fig. 10. immunocytochemical demonstration of fibronectin along the cell surface and cell processes of mesoblast cells (× 20340). Fig. 11. Fluorescence photomicrograph of the apical surface of the upper layer stained in toto with FITC-conjugated phalloidin. The picture shows the presence of polygonal networks composed of actin filaments of the zonulae adhaerentes ( × 825). Fig. 12. Transmission electron photomicrograph of an upper-layer cell at the level of the primitive streak, showing the alignment of microtubules (arrows) along the apicobasal axis of the cell ( × 22150). Fig. 13. Scanning electron photomicrograph of a migrating mesoblast cell extending a lamellipodium and filopodia on the basement membrane of the epiblast (UL), which is used as a substrate ( × 1850). Fig. 14. Photomicrograph of a mesoblast cell in the vicinity of the band of fibrils situated in the endophyllic crescent, at the edge of the area pellucida. The cells appear more spherical, and they have retracted their cell processes ( × 2540). 96 cAMP, causing the so-called axial bending (Robertson et al., 1978; Gingle and Robertson, 1979). Treatment of the embryo with external cAMP phosphodiesterase, a cAMP-degrading enzyme that colocalizes with microfilament bundles (Neuman et al., 1983), stops the gastrulation (Robertson and Gingle, 1977). Another evidence for the involvement of cAMP in ingression of cells is the presence of adenylate cyclase activity according to a gradient decreasing from the primitive streak towards the lateral edge of the area pellucida (Neuman et al., 1983): Increased amounts of cAMP in the region of ingression of upper-layer cells have, however, not been confirmed ultrastructurally by Sanders (1987). Finally, it has been suggested that serotonin may be another important component that may primarily promote the activity of microfilaments and microtubules, since treatment with this compound interferes with the formation of the primitive streak (Pal~n et al., 1979). The contention that microfilaments bundles are the primary forces driving ingression of upperlayer cells has partly been based on the use of drugs that disrupt microfilaments. Cytochalasin B causes middle-layer and upper-layer cells to round up, and middle-layer cells withdraw from the substrate provided by the basement membrane of the upper layer (De-Voy et al., 1979). A similar effect has been noted in deep-layer cells, which dissociate. Treatment results in rounded cells devoid of microappendages and covered by large blebs (Stagno and Low, 1978; Chamorro et al., 1986). Although cytochalasin B has no effect on the cell shape of the cells in the margin of overgrowth. suggesting that migration of these cells does not require major changes in the distribution of microfilaments (Chernoff and Overton, 1979), a number of reports have indicated that morphogenetic processes cease in the presence of this drug (Karfunkel, 1972; Lee and Kalmus, 1976; Chernoff and Lash, 1981; Ostrovsky et al., 1983). One has, however, to keep in mind that the effect of cytochalasin B on cell shape, that is usually attributed to a disruptive effect on actin microfilaments, may be obtained through side effects on the cell surface (see Nardi, 1981). As an example, cytochalasin B releases fibronectin from cell mem- branes (Kurkinen et al., 1978). Microtubules have also been localized in upper-layer cells, and they are arranged parallel to the apicobasal axis of the cells (Fig. 12) (Balinsky and Walther, 1961; Granholm and Baker, 1970: Sanders and Zalik, 1970; Bancroft and Bellairs, 1975). Their importance in the maintenance of the elongate shape has been suggested on the basis of experiments with colchicine, a microtubule inhibiting agent that causes the cells of the primitive streak to round up (Granholm, 1970). Assuming that the integrity of the microtubules is essential for directional migration, Mareel et al. (1984) investigated the role of the microtubule inhibitors, nocodazole and taxol, in the expansion of fragments of hypoblast and of the margin of overgrowth in culture. They observed an arrest of cell spreading in both cases, suggesting that the integrity of the microtubular complex is another prerequisite for epiboly and migration of hypoblast cells. From the results reviewed here, it seems obvious that cytoskeletal elements are implicated in early morphogenetic events. Whether their behaviour is influenced by transmembrane signals from the extracellular matrix is not clear. Indirect evidence for this assumption has, however, been provided by Vanroelen et al. (1982), who observed that the removal of the deep layer induces changes in the structure of the basement membrane of the wounded area. These changes are correlated with the appearance of cell junctions and of microfilament bundles at the basal side of the upper-layer cells underlain by an altered basement membrane. Cell-surface specializations The transfer of chemical messengers from one cell to another as mediator for short-range intercellular communication implies physical contact via cell junctions or close apposition of cell surfaces. The apposition of cell surfaces may occur through the presence of microappendages of the cell, and their presence, as well as the chemical properties of the cell surface, may be an essential factor in inductive processes and in control of cell behaviour. As an example, Eyal-Giladi and Wolk 97 (1970) have shown that the induction of a primitive streak by the hypoblast is dependent on direct cellular contact between the two layers. In this paragraph, we shall describe the distribution and the nature of the cell microappendages that are encountered during gastrulation. The morphology of these structures is, however, largely dependent on the fixation procedure (Litke and Low, 1977; Litke, 1980; Andries and Vakaet, 1985a), and the variability of this parameter is probably the origin of several discrepancies noted in the literature. Regional differences in cell morphology have been found in the upper layer from early gastrulation on. Indeed, the cells of the central area are relatively thicker than the cells of the periphery of the area pellucida (Vakaet, 1970) and, using scanning electron microscopy, Bancroft and Bellairs (1974) and Jacob et al. (1974) have shown that the former cells, which are destined to ingress, possess fewer apical microvilli than the cells of lateral regions. Cilia and connecting cords (beaded threads) have also been described (Jacob et al., 1974; Bancroft and Bellairs, 1975). Dimples and crypts (Bancroft and Bellairs, 1974), and folds and pits (Weinberger and Brick, 1982a) have been described along the apical surface of the area pellucida of the chicken embryo, but were not found by Andries et al. (1983a) in just laid quail blastoderms, with exception of a few dimple-like depressions. Until laying, the cells of the area opaca are larger than the cells of the area pellucida (Kochav et al., 1980), and they possess less microvilli, which are mainly found along the cell borders (Andries et al., 1983a). Finally, the cells of the margin of overgrowth extend iamellipodia and filopodia, suggesting their role in locomotion (Andries et al., 1983a). The ventral surface of the upper layer is characterized by the presence of a basement membrane that is interrupted at the level of the primitive streak. Few surface specializations have been noted, except along bottleshaped cells of the primitive-streak region, which share filopodia at the leading edge (Bancroft and Bellairs, 1975), broad flat processes (Wakely and England, 1977), and numerous blebs (Vakaet et al., 1980), and along upper-layer cells destined to incorporate into the deep layer by polyingression (Weinberger et al., 1984). During the onset of formation of the hypoblast - the 'sickle endoblast' of Vakaet (1970) and the 'primary hypoblast' of Weinberger and Brick (1982a,b) - cells of the upper layer drop down onto the dorsal surface of the deep layer, from stage 1 on, in the posterior part of the blastoderm, and incorporate into this layer to form the hypoblast. These cells show extensive microprojections, and intercellular contacts are made (Sanders et al., 1978; Weinberger and Brick, 1982a). Blebbing and local interruptions in the basement membrane are also evident. At the level of the area opaca, the subgerminal yolk surface is covered by microvilli and small pits (Andries et al., 1983b) during stages X-XIV of Eyal-Giladi and Kochav (1976). More laterally, the ventral cells of the germ wall are attached to the subgerminal yolk by filopodial extensions. From stage XIII on, these cells are usually more flattened and they protrude lamellipodia and filopodia (Andries et al., 1983b). The cells of the primitive streak destined to differentiate into primary mesenchyme become stellate-shaped soon after their emigration from the primitive streak. They possess long, narrow filopodia on the leading edge of the cells and, to some extent, on the lateral surfaces (Hay, 1968; Litke, 1978b). Lamellipodia and a large number of filopodia reach the basement membrane of the epiblast, which is used by these cells as a substrate for migration (Fig. 13) (Revel and Solursh, 1978, 1986). Thread-like extensions, probably retraction fibres, have been found on the trailing edge of migrating cells. However, in the vicinity of the band of extracellular fibrils situated beneath the upper layer at the anterior and lateral borders of the area pellucida (see England, 1981, 1982), the cells are spherical and possess blebs and retraction fibres only (Fig. 14). These changes in cell morphology have been related to the behaviour of the mesoblast cells (Andries et al., 1985b). The deep layer is built-up by tissues of different origin and fate. Apart from the observation that the surface morphology of cells is dependent on fixation, the fact that authors have used different terms, to refer to the whole deep layer at all stages, obstructs comparisons between different reports. We shall restrict this paragraph to welldefined observations reported several times. The 98 t O .Q _ t ~-, . e" 1 Figs. 15-18. Scanning electron photomicrographs on the ventral surface of the different types of cells composing the deep layer. A distinction between endophyll cells (Fig. 15, x5900), definitive-endoblast cells (Fig. 16, × 1700), hypoblast cells (Fig. 17, × 1700), and marginal (junctional) hypoblast cells (Fig. 18, × 5300) is made on the basis of surface morphology. identification of the different tissues on the basis of surface specializations at the dorsal surface is difficult, because most of the cells possess long microvilli reaching the cells of the middle and upper layers. The cells of the definitive endoblast are more closely apposed to one another than the other cells of the deep layer (Vakaet and Hertoghs-De Maere, 1973; Sanders et al., 1978; Stolinsky et al., 1981). The hypoblast has large holes in it, and when it is viewed from the ventral side, one can see the middle layer through the holes (Revel, 1974). Microappendages of mesoblast cells protrude through these holes in the hypoblast and reach the subgerminal cavity (Low et al., 1975; Litke, 1978b; Highison and Low, 1981). From the study of the ventral surface, facing to the yolk, regional differences appear (Sanders et al., 1978; Weinberger and Brick 1982a,b; Bellairs, 1982, for review, Andries et al., 1987). An extreme pleomorphism among the microappendages (microvilli, blebs, ruffles) has been recognized by Harri and Low (1974), but no • 99 clear-cut arrangement was found. During mitosis, however, the cells of the deep layer round up, and the number of microvilli increases until late metaphase or anaphase. Ruffles and blebs are not prominent, and filopodia are absent (Harri and Low, 1975). Microappendages, especially ruffles, have also been found in contact with yolk granules, suggesting their role in phagocytosis (Litke and Low, 1975). During the onset of hypoblast formation, the number of microappendages is low, whereas these structures are numerous prior to and following hypoblast formation (Litke, 1978a). Regional differences appear more obviously from results of Litke (1978b), who reported that cells of the periphery of the area pellucida are covered by ruffles and microvilli, whereas in the caudal region the cells are larger and associated with microplicae. These results may be considered in line with these of Wakely and England (1978), showing that cells delaminating from the 'ectoderm' (i.e. those forming hypoblast) are round and virtually free of cell processes, but become covered by microvilli afterwards. Moreover, the cells originating from the primitive streak (i.e. the definitive endoblast) are either smooth or have short microvilli after their insertion into the deep layer. Observations made in this laboratory confirm these views (Andries et al., 1987). The ventral surface of the chicken blastoderm may be divided in different zones, according to their surface morphology: the endophyllic crescent is covered by numerous microvilli (Fig. 15); the definitive endoblast is relatively smooth, with few microvilli and some cilia protruding in the subgerminal cavity (Fig. 16); the hypoblast, which is composed of a loose network of cells, possess somewhat more microvilli (Fig. 17). The posterior part of the hypoblast, referred to as junctional endoblast (Vakaet, 1970; Sanders et al., 1978), and the marginal hypoblast, at the edge of the area pellucida, are covered with microplicae (Fig. 18). From stage 6 on, microvilli have been observed as well. Finally, the yolk endoderm of the area opaca shares numerous ruffles. The ventral surface of the deep layer is responsive to experimental manipulation including the composition of the fixative (Litke and Low, 1977; Litke, 1980) as indicated earlier, but also to treatment with cytochalasin B (Stagno and Low, 1978), coichicine (Harriet al., 1978), anesthetics (Stagno and Low, 1980), and insulin-glucose mixtures (Highison and Low, 1981). Apart from the cell microappendages, most of the cell membranes contain domains specialized in receptor-mediated endocytosis and known as coated pits (Pastan and Willingham, 1983 for review). The presence of coated pits in the area pellucida (Raveh et al., 1971) and their implication in active and selective uptake of yolk into the epiblast (MacLean and Sanders, 1983) have been reported. Unexpected is the observation that, at the ventral side of the deep layer, these structures are associated with fibronectin (Sanders, 1982; Harrisson et al., 1985c). This extensive overview of the distribution and nature of cell-surface specializations has documented the distinctness of the different tissues, and the changes in protrusive activity that occur during the period of gastrulation. Discussing the role of protrusive activity of the cell surface in the initiation of morphogenetic cell movements would be irrelevant and misplaced after the excellent review that has been published by Trinkaus (1985) on this subject. Electrical fields The role of ionic currents in the establishment of developmental pattern has been reviewed by Jaffe (1981), and a model for early chicken morphogenesis, based on the existence of electrical currents, has been proposed by Stern (1984). During gastrulation in the chicken, extracellular electrical current pathways, indeed, flow out of the dorsal region of the primitive streak, and return to the epiblast via the primitive streak in central areas of the blastoderm (Jaffe and Stern, 1979). Using autoradiographic labelling with ouabain, which labels specifically the sodium/potassium pump, Stern (1982a) and Stern and MacKenzie (1983) demonstrated labelling along the basal surface of the upper-layer cells that are not underlain by mesoblast cells. In the primitive-streak region and in regions colonized by mesoblast cells, the pump is localized at the apical surface of the upper layer. These results suggest that the currents originate from a pump operating in the epiblast, 100 accumulating sodium and water between the germ layers. The pattern of extracellular current flow is opposite to the direction of movement of cells in the upper and middle layers. To test the hypothesis of a causal relationship between these two phenomena, Stern (1981) placed isolated middle-layer cells in different electrical fields. His results demonstrate that the presence of electrical currents flowing in the extracellular space do not play a direct role in the orientation of migration of single middle-layer cells. In the embryo, however, electrical fields may polarize substrate-attached material. This polarization is lost if serum is added or after trypsin digestion, suggesting that a serum-derived protein, such as fibronectin, may play this role (see Sanders, 1980, and Stern, 1981). Another role for electrical currents may be the maintenance of cell polarity. This hypothesis results from the observation that the application of voltages of similar magnitude, but of reverse polarity to the normal potentials, across the upper layer causes a reversal of polarity of the cells (Stern, 1982b; Stern and MacKenzie, 1983), since apical alcian-blue-stainable extracellular materials and basal tight junctions have been found. It thus appears that sodium transport, reflected by electrical currents, is an important physiological function that primarily determines the polarity of the upper layer, the polarization of extracellular materials and, consequently, the direction of migration of tissues. Concluding remarks Examination of the abundant literature on chicken gastrulation reveals that the interest that has been focussed for a long time on cellular aspects of inductive processes guiding morphogenesis has now been shifted to a different level: under the impulse of cell biological studies and of experimental approaches, research has witnessed some fundamental changes. Research has, indeed, focussed increasingly on the molecular aspects of cell and tissue interactions and on their role in overall cell functions. In this review, we have attempted to collect and integrate some of the major currents in the study of intercellular communication in the gastrulating chicken embryo, viewed insofar as possible from the standpoint of cell biology. In the last decade, the biological examination of the phenomenon of intercellular communication in morphogenesis has allowed a considerable increase in the understanding of morphogenetic events, but it often led to a conceptual isolation of different aspects from their surroundings. The major theme that we have tried to put forward in this review has been that several cell biological aspects of intercellular communication probably all have to operate simultaneously to obtain changes in cell shape and cell adhesion that, in turn, lead to the positioning of the different germ layers. As an indication of the obligatory cooperation of these several processes, the initiation of migration of mesoblast cells requires changes in the composition of the extracellular matrix, loss of junctional communication between the cells, redistribution of cytoskeletal elements responsible for changes in cell shape, development of cell-surface specializations responsible for migration on a substrate, and presence of adequate extraceilular currents responsible for polarization of extracellular material and maintenance of cell polarity. Although considerable progress has been made during the past years towards understanding the cell biological aspects guiding developmental events, many important questions remain unanswered. For example, what are the factors governing the arrest of cell migration at a given place? How can we discriminate between causes and effects of migration? Is the expression of extracellular compounds necessarily related to developmental events or is it circumstantial and should it be regarded as a marker of differentiation? Which cells of the upper layer are committed to form the definitive endoblast or the mesoblast? At what moment are the cell commitments established? The application of new techniques should certainly bring new and original insights into these problems and assist us in clarifying cell and developmental processes. Acknowledgements The authors wish to express their gratitude to their sometime collaborators, Dr. J. Van Hoof and Dr. Ch. Vanroelen, for advice and criticism, to 101 Mr. F. De Bruyn, Mr. Ch. De Schepper, Mrs. V. Gonthier-Van der Stock, Mrs. E. Haest-Van Nueten, and Mr. G. Van den Broeck for excellent technical assistance, to Mr. J. Van Ermengem for photographic processing, and to Mrs. N. Van den Hende-Bol for typing the manuscript. References Abrahamsohn. P.A., J.W. Lash, R.A. Kosher and RR. Minor: The uniquitous occurrence of chondroitin sulfates in chick embryos. J. Exp. Zool. 194, 511-518 (1975). Andries. L., F. Harrisson and M. Callebaut: Differential distribution of cell protrusions on the ventral surface of the deep layer in gastrulating quail and chick embryos. Differentiation 34, 168-174 (1987). Andries, L., F. Harrisson, R. Hertsens and L. Vakaet: Cell junctions and locomotion of the blastoderm edge in gastrulating chick and quail embryos. J. Cell Sci, 78, 191-204 (1985a). Andries, L. and L. Vakaet: The morphology of the migrating epiblast edge in 8astrulating quail embryos: an SEM study of the effects of tissue handling before fixation on the attachment side. J. Exp. Zool. 236, 181-191 (1985a). Andries, L. and L. Vakaet: Cell junctions in the rapidly moving edge of chick and quail blastoderms during gastrulation. Cell Tissue Res. 242, 157-162 (1985b). Andries, L., L. Vakaet and Ch. Vanroelen: The dorsal surface of the animal pole of the just laid quail egg, studied with SEM. Anat. Embryol. 166, 135-147 (1983a), Andries, L., L. Vakaet and Ch. Vanroelen: The subgerminal yolk surface and its relationship with the inner germ wall edge of the stages X to XIV chick and quail embryos. Anat. Embryol. 166, 453-462 (1983b). Andries, L., Ch. Vanroelen, J. Van Hoof and L. Vakaet: Inhibition of cell spreading on the band of extracelhilar fibres in early chick and quail embryos. J. Cell Sci. 74, 37-50 (1985b). Balinsky. B.I. and H. Walther: The immigration of the presumptive mesoblast from the primitive streak in the chick as studied with the electron microscope. Acta Embryol. Morphol. Exp. (Palermo) 4, 261-283 (1961). Bancroft. M. and R. BeUairs: The onset of differentiation in the epiblast of the chick blastoderm (SEM and TEM). Cell Tissue Res. 155, 399-418 (1974). Bancroft, M. and. R. Bellairs: Differentiation of the neural plate and neural tube in the young chick embryo. Anat. Embryol. 147, 309-335 (1975). Bellairs, R.: Differentiation of the yolk sac of the chick studied by electron microscopy. J. Embryol. Exp. Morphol. 11, 201-225 (1963). Bellairs, R.: Developmental Processes in Higher Vertebrates (Logos Press Ltd., London) 366 pp. (1971). Bellairs, R.: Gastrulation processes in the chick embryo. In: Cell Behaviour. A Tribute to Michael Abercrombie, eds. R. Bellairs, A. Curtis and G. Dunn (Cambridge University Press, Cambridge) pp. 395-427 (1982). 13ellairs, R.: The primitive streak. Anat. Embryol. 174, 1-14 (1986). Bellairs, R., A.S. Breathnach and M. Gross: Freeze-fracture replication of junctional complexes in unincubated and incubated chick embryos. Cell Tissue Res. 162, 235-252 (1975). Bellairs, R. and D.A.T. New: Phagocytosis in the chick blastoderm. Exp. Cell Res. 26, 275-279 (1962). Chamorro, C.A., P. de Paz Cabello, J.G. Fernandez and J.M. Villar: Effects of cytochalasin B on cell to cell adhesion and cellular shape of chick embryo mesoderm cells in vitro. J. Anat. 147, 201-206 (1986). Chernoff, E.A.G. and J.W. Lash: Cell movement in somite formation and development in the chick: Inhibition of segmentation. Dev. Biol. 87, 288-298 (1981). Chernoff, E.A.G. and J. Overton: Organization of the migrating chick epiblast edge: attachment sites, cytoskeleton, and early developmental changes. Dev. Biol. 72. 291-307 (1979). Clarke, M. and J.A. Spudich: Nonmuscle contractile proteins: The role of actin and myosin in cell motility and shape determination. Annu. Rev. Biochem. 46, 797-822 (1977). Critchley, D.R., M.A. England, J. Wakely and R.O. Hynes: Distribution of fibronectin in the ectoderm of gastrulating chick embryos. Nature 280, 498-500 (1979). De-Voy. K., M.A. England and J. Wakely: The effect of cytochalasin B on chick mesoderm cells as studied by scanning electron microscopy. Anat. Embryol. 158, 63-73 (1979). Douglas Powers, R. and J.T. Tupper: Intercellular communication in the early embryo. In: Intercellular Communication, ed. W.C. De Mello (Plenum Press. New York) pp. 231-251 (1977). Downie, J.R. and S.M. Pegrum: Organization of the chick blastoderm edge. J. Embryol. Exp. Morphol. 26, 623-635 (1971). Duband, J.L. and J.P. Thiery: Appearance and distribution of fibronectin during chick embryo gastrulation and neurulation. Dev. Biol. 94, 337-350 (1982). Ebendal, T.: Migratory mesoblast cells in the young chick embryo examined by scanning electron microscopy. Zoon 4, 101-108 (1976). Edelman, G.M., W.J. Gallin, A. Delouvee, B.A. Cunningham and J.P. Thiery: Early epochal maps of two different cell adhesion molecules. Proc. Natl. Acad. Sci. USA 80, 4384-4388 0983). England, M.A.: Applications of the SEM to the analysis of morphogenetic events. J. Microsc. 123, 133-146 (1981). England, M.A.: Interactions at basement membranes. In: The Functional Integration of Cells in Animal Tissues. eds. J.D. Pitts and M.E. Finbow (Cambridge University Press, Cambridge) pp. 209-228 (1982). Eyal-Giladi, H,: The gradual establishment of cell commitments during the early stages of chick development. Cell Differ. 14, 245-255 (1984). Eyal-Giladi, H. and S. Kochav: From cleavage to primitive streak formation: A complementary normal table and a 102 new look at the first stages of the development of the chick. I. General morphology. Dev. Biol. 49, 321-337 (1976). EyaI-Giladi, H. and M. Wolk: The inducing capacities of the primary hypoblast as revealed by transfilter induction studies. Roux's Arch. Dev. Biol. 165, 226-241 (1970). Fisher, M. and M. Solursh: Glycosaminoglycan localization and role in maintenance of tissue spaces in the early chick embryo. J. Embryol. Exp. Morphol. 42, 195-207 (1977). Fisher, M. and M. Solursh: The influence of local environment on the organization of mesenchyme cells. J. Embryol. Exp. Morphol. 49, 295-306 (1979a). Fisher, M. and M. Solursh: The influence of the substratum on mesenchyme spreading in vitro. Exp. Cell Res. 123, 1-14 (1979b). Gingle, R. and A. Robertson: Response of the early chick embryo to external cAMP sources. J. Embryol. Exp. Morphol. 53, 353-365 (1979). Goodwin, B.G. and M.H. Cohen: A phase-shift model for the spatial and temporal organization of developing systems. J. Theor. Biol. 25, 49-107 (1969). Granholm, N.H.: Effects of microtubular inhibiting agents on cell shape in the primitive streak of the avian embryo. Am. Zool. 10, 320 (1970). Granholm, N.H. and J.R. Baker: Cytoplasmic microtubules and the mechanism of avian gastrulation. Dev. Biol. 23, 563-583, (1970). Hamburger, V. and H. Hamilton: A series of normal stages in the development of the chick embryo. J. Morphol. 88, 49-92 (1951). Harri, J.E. and F.N. Low: Microappendages on the ventral surface of the primitive streak chick embryo. Am. J. Anat. 141,569-573, (1974). Harri, J.E. and F.N. Low: Mitotic cells and their microappendages in the primitive streak of the chick embryo. Am. J. Anat. 144, 249-254 (1975). Harri, J.E., P.A. Stagno and F.N. Low: Scanning electron microscopy of the ventral surface of early chick embryos treated with colchicine. J. Submicrosc. Cytol. 10. 299-308 (1978). Harrisson, F.: Autoradiographic study of the origin of basement membrane components in the avian embryo. In: Progress in Developmental Biology, Part B, ed. H.C. Slavkin (Alan R. Liss, New York), pp. 377-380 (1986). Harrisson, F., J. Van Hoof and J.M. Foidart: Demonstration of the interaction between glycosaminoglycans and fibronectin in the basement membrane of the living chicken embryo. Roux's Arch. Dev. Biol. 193, 418-421 (1984a). Harrisson, F., J. Van Hoof, Ch. Vanroelen and J.M. Foidart: Masking of antigenic sites of fibronectin by glycosaminoglycans in ethanol-fixed embryonic tissue. Histochemistry 82, 169-174 (1985a). Harrisson, F., J. Van Hoof, Ch. Vanroelen and L. Vakaet: Transfer of extracellular matrix components between germ layers in chimaeric chicken-quail blastoderms. Cell Tissue Res. 239, 643-649 (1985b). Harrisson, F., Ch. Vanroelen, J.M. Foidart and L. Vakaet: Expression of different regional patterns of fibronectin immunoreactivity during mesoblast formation in the chick blastoderm. Dev. Biol. 101, 373-381 (1984). Harrisson, F., Ch. Vanroelen and I_. Vakaet: Fibronectin and its relation to the basal lamina and to the cell surface in the chicken blastoderm. Cell Tissue Res. 241,391-397 (1985c). Hay, E.D.: Organization and fine structure of epithelium and mesenchyme in the developing chick embryo. In: Epithelial-Mesenchymal Interactions, eds, R. Fleischmajer and R.E. Billingham (The Williams & Wilkins Company. Baltimore) pp. 31-63 (1968). Highison, G.J. and F.N. Low: The effect of insulin-glucose on microappendages of the plasmalemma in the early chick embryo. Am. J. Anat. 161,405-413 (1981). Ireland, G.W., E.J. Sanders. F.C.T. Voon and .1. Wakely: The ultrastructure of polygonal networks in chick embryonic cells in vitro. Cell Biol. Int. Rep. 7, 679-688 (1983). Ireland, G.W. and F.C.T. Voon: Polygonal networks in living chick embryonic cells. J. Cell. Sci. 52. 55-69 (1981). Jacob, H.J., B. Christ, M. Jacob and G.J. Bijvank: Scanning electron microscope (SEM) studies on the epiblast of young chick embryos. Z. Anat. Entwicklsgesch. 143, 205-214 (1974). Jaffe, L.F.: The role of ionic currents in establishing developmental patterns. Philos. Trans. R. Soc. Lond. B 295. 553-566 (1981). Jaffe, L.F. and C.D. Stern: Strong electrical currents leave the primitive streak of chick embryos. Science 206, 569-571 (1979). Karfunkel, P.: The activity of microtubules and microfilaments in neurulation in the chick. J. Exp. Zool. 181. 289-302 (1972). Kemler, R., D. Morello, Ch. Babinet and F. Jacob: Surface antigens in early embryonic development. In: Differentiation and Neoplasia, eds. R.G. McKinnell, M.A. De Berardino, M. Blumenfeld and R.D. Bergad (Springer-Verlag, Berlin) pp. 107-111 (1980). Kochav, S., M. Ginsburg and H. EyaI-Giladi: From cleavage to primitive streak formation: a complementary normal table and a new look at the first stages of the development of the chick. Ii. Microscopic anatomy and cell population dynamics. Dev. Biol. 79, 269-308 (1980). Koller, C.: Untersuchungen iiber die Bl~itterbildung im Hiihnerkeim. Arch. Mikr. Anat. 20, 174-211 (1882). Krotoski, D.M., C. Domingo and M. Bronner-Fraser: Distribution of a putative cell surface receptor for fibronectin and laminin in the avian embryo. J. Cell Biol. 103, 1061-1071 (1986). Kucera, P. and M.B. Burnand: Mechanical tension and movement in the chicken blastoderm as studied by real-time image analysis. J. Exp. Zool. Suppl. 1, 329-339 (1987). Kurkinen, M., J. Wartiovaara and A. Vaheri: Cytochalasin B releases a major surface-associated glycoprotein, fibronectin, from cultured fibroblasts. Exp. Cell Res. 111, 127-137 (1978). Le Douarin, N.M.: Cell recognition based on natural morphological nuclear markers. Med. Biol. 52, 281-319 (1974). Lee, H.Y. and G.W. Kalmus: Effects of cytochalasin B on the morphogenesis of explanted early chick embryos. Growth 40, 153-162 (1976). 103 Litke, L.L.: Development and hypoblast formation in early chick embryos as observed by scanning electron microscopy. J. Submicrosc. Cytol. 10, 1-13 (1978a). Litke, L.L.: The ventral surface of the early chick embryo: A scanning electron microscopic study. J. Submicrosc. Cytol. 10, 287-298 (1978b). Litke, L.L.: The effects of various experimental substrates on the ventral surface of early chick embryos as observed by electron microscopy. Micron 11, 11-12 (1980). Litke, L.L. and F.N. Low: Scanning electron microscopy of yolk absorption in early chick embryos. Am. J. Anat. 142, 527-532 (1975). Litke, L.L. and F.N. Low: Fixative tonicity for scanning electron microscopy of delicate chick embryos. Am. J. Anat. 148, 121-127 (1977). Low, F.N., J.E. Harri and L.L. Litke: The early chick embryo in correlated SEM and TEM studies. Scan. Electron Microsc. 1, 403-410 (1975). Macarak, E.J.: An ultrastructural study of adhesive junctions in reaggregates of unincubated chick embryos. J. Morphol. 147, 41-60 (1976). MacLean, I.M. and E.J. Sanders: Cationized ferritin and phosvitin uptake by coated vesicles of the early chick embryo. Anat. Embryol. 166, 385-397 (1983). Manasek, F.J.: The extracellular matrix: a dynamic component of the developing embryo. In: Current Topics in Developmental Biology, eds. A.A. Moscona and A. Monroy (Academic Press, New York) pp. 35-102 (1975). Marcel, M., R. Bellairs, G. De Bruyne and M.R. Van Peteghem: Effect of microtubule inhibitors on the expansion of hypoblast and margin of overgrowth in chick embryos. J. Embryol. Exp. Morphol. 81,273-286 (1984). Marcel, M. and L. De Ridder: The margin of overgrowth of the embryonic chick blastoderm as a study model for cell-to-cell contacts. Experientia 26, 1003-1005 (1970). Mitrani, E.: Primitive streak-forming cells of the chick invaginate through a basement membrane. Roux's Arch. Dev. Biol. 191,320-324 (1982). Mitrani, E. and A. Farberov: Fibronectin expression during the processes leading to axis formation in the chick embryo. Dev. Biol. 91, 197-201 (1982). Monnet-Tschudi, F., P. Favrod, M.B. Burnand, C. Verdan and P. Kucera: Fibronectin in the area opaca of the young chick embryo, lmmunofluorescence and immuno-electronmicroscopic study. Cell Tissue Res. 241, 85-92 (1985). Moscona, A.A.: Surface specification of embryonic cells: lectin receptors, cell recognition, and specific cell ligands. In: The Cell Surface in Development, ed. A.A. Moscona (John Wiley & Sons, New York) pp. 67-99 (1974). Nardi, J.B.: Epithelial invagination: adhesive properties of cells can govern position and directionality of epithelial folding. Differentiation 20, 97-103 (1981). Neuman, T., T. Laasberg and J. K~irner: Ultrastructurai localization of adenylate cyclase and cAMP phosphodiesterase in the gastrulating chick embryo. Roux's Arch. Dev. Biol. 192, 42-44 (1983). New, D.A.T.: The adhesive properties and expansion of the chick blastoderm. J. Embryol. Exp. Morphol. 7, 146-164 (1959). Ostrovsky, D., J.W. Sanger and J.W. Lash: Light microscope observations on actin distribution during morphogenetic movements in the chick embryo. J. Embryol. Exp. Morphol. 78, 23-32 (1983). Overton, J.: Desmosome development in normal and reassociating cells in the early chick blastoderm. Dev. Biol. 4, 532-548 (1962). Overton, J.: Experiments with junctions of the adhaerens type. In: Current Topics in Developmental Biology, eds. A.A. Moscona and A. Monroy (Academic Press, New York) pp. 1-34 (1975). Pal~n, K., L. Tht~rneby and H. Emanuelsson: Effects of serotonin and serotonin antagonists on chick embryogenesis. Roux's Arch. Dev. Biol. 187, 89-103 (1979). Pastan, I. and M.C. Willingham: Receptor-mediated endocytosis: coated pits, receptosomes and the Golgi. Trends Biochem. Sci. 8, 250-254 (1983). Raveh, D., M. Friedlander and H. EyaI-Giladi: Organelle differentiation in the chick blastoderm during hypoblast formation. Wilhelm Roux Arch. Entw. Mech. Org. 166, 287-299 (1971). Revel, J.P.: Some aspects of cellular interactions in development. In: The Cell Surface in Development, ed. A.A. Moscona (John Wiley & Sons, New York) pp. 51-65 (1974). Revel, J.P. and M. Solursh: Ultrastructure of primary mesenchyme in chick and rat embryos. Scan. Electron Microsc. 11, 1040-1046 (1978). Revel, J.P. and M. Solursh: Uhrastructure of primary mesenchyme in chick and rat embryos. In: Scanning electron microscopy studies of embryogenesis, ed. G.C. Schoenwolf (Scanning Electron Microscopy Inc., Chicago) pp. 75-80 (1986). Revel, J.P., P. Yip and L.L. Chang: Cell junctions in the early chick embryo. A freeze etch study. Dev. Biol. 35, 302-317 (1973). Robertson, A. and A.R. Gingle: Axial bending in the early chick embryo by a cyclic adenosine monophosphate source. Science 197, 1078-1079 (1977). Robertson, A., J.F. Grutsch and A.R. Gingle: Cyclic adenosine monophosphate production by embryonic chick ceils. Science 199, 9900991 (1978). Sanders, E.J.: Intercellular contact in the unincubated chick embryo. Z. Zellforsch. 141,459-468 (1973). Sanders, E.J.: Development of the basal lamina and extracellular materials in the early chick embryo. Cell. Tissue Res. 198, 527-537 (1979). Sanders, E.J.: The effect of fibronectin and substratum-attached material on the spreading of chick embryo mesoderm cells in vitro. J. Cell. Sci. 44, 225-242 (1980). Sanders, E.J.: Ultrastructural immunocytochemical localization of fibronectin in the early chick embryo. J. Embryol. Exp. Morphol. 71,155-170 (1982). Sanders, E.J.: Recent progress towards understanding the roles of the basement membrane in development. Can. J. Biochem. Cell. Biol. 61,949-956 (1983). Sanders, E.J.: Mesoderm migration in the early chick embryo. In: Developmental Biology. A Comprehensive Synthesis, 104 Vol. 2, ed. L. Browder (Plenum Pub. Corp., New York) pp. 449-480 (1986a). Sanders, E.J.: Cytochemistry of the cell surface and extracellular matrix during early embryonic development. Prog. Histochem. Cytochem. 16, 1-57 (1986b). Sanders, E.J.: Ultrastructural cytochemical localization of adenylate cyclase in the early chick embryo. Cell. Tissue Res. 247, 465-468 (1987). Sanders, E.J., R. Bellairs and P.A. Portch. In vivo and in vitro studies on the hypoblast and definitive endoblast of avian embryos. J. Embryol. Exp. Morphol. 46, 187-205 (1978). Sanders, E.J. and S.E. Zalik: The occurrence of microtubules in the pre-streak chick embryo. Protoplasma 71, 203-208 (1970). Sax6n, L.: Morphogenetic tissue interactions: An introduction. In: Cell Interactions in Development, eds. M. KarkinenJ~i~iskel~iinen. L. Sax6n and L. Weiss (Acaderruc Press, New York) pp. 145-151 (1977). Saxdn, L.: Mechanism of morphogenetic tissue interactions: The message of transfilter experiments. In: Differentiation and Neoplasia, eds. R.G. McKinnell, M.A. Di Berardino, M. Blumenfeld and R.D. Bergad (Springer-Verlag, Berlin) pp. 147-154 (1980). Slavkin, H.C.: Progress in Clinical and Biological Research, Vol. 217, Parts A and B: Progress in Developmental Biology (Alan R. Liss, New York) 448 pp. and 462 pp. (1986). Solursh, M.: Glycosaminoglycan synthesis in the chick gastrula. Dev. Biol. 50, 525-530 (1976). Solursh, M. and J.P. Revel: A scanning electron microscope study of cell shape and cell appendages in the primitive streak region of the rat and chick embryo. Differentiation 11, 185-190 (1978). Stagno, P.A. and F.N. Low: Effects of cytochalasin B on the fine structure of organized endodermal cells of the early chick embryo. Am. J. Anat. 151, 159-172 (1978). Stagno, P.A. and F.N. Low: Anesthetic action on the fine structure of the plasmalemma and junctional complexes of chick endodermal cells. J. Submicrosc. Cytol. 12, 321-344 (1980). Stem, C.D.: Behaviour and motility of cultured chick mesoderm cells in steady electrical fields. Exp. Cell. Res. 136, 343-350 (1981). Stern, C.D.: Localisation of the sodium pump in the epiblast of the early chick embryo. J. Anat. 134, 606-607 (1982a). Stem, C.D.: Experimental reversal of polarity in chick embryo epiblast sheets in vitro. Exp. Cell Res. 140, 468-471 (1982b). Stern, C.D.: A simple model for early morphogenesis. J. Theor. Biol. 107, 229-242 (1984). Stern, C.D. and D.O. MacKenzie: Sodium transport and the control of epiblast polarity in the early chick embryo. J. Embryol. Exp. Morphol. 77, 73-98 (1983). Stolinski, C., E.J. Sanders. R. Bellairs and B. Martin: Cell junctions in explanted tissues from early chick embryos. Cell. Tissue Res. 221,395-404 (1981). Thiery, J.P.: Fibronectin in early embryonic development of the vertebrate. Bio Essays 2, 32-34 (1985). Thiery, J.P.. A. Delouv6e, W.J. Gallin, B.A. Cunningham and G.M. Edelman: Ontogenic expression of cell adhesion molecules: L-CAM is found in epithelia derived from the three primary germ layers. Dev.Biol. 102. 61-78 (1984). Thiery, J.P., J.L. Duband, V. Rutishauser and G.M. Edelman: Cell adhesion molecules in early chicken embryogenesis. Proc. Natl. Acad. Sci. USA 79, 6737-6741 (1982). Toole, B.P.: Developmental role of hyaluronate. Connect. Tissue Res. 10, 93-100 (1982). Trelstad, R.L., E.D Hay and J.P. Revel: Cell contact during early morphogenesis in the chick embryo. Dev. Biol. 16, 78-106 (1967). Trelstad, R.D., J.P. Revel and E.D. Hay: Tight junctions between cells in the early chick embryo as visualized with the electron microscope. J. Ceil. Biol. 31, C6-C10 (1966). Trinkaus, J.P.: Cells into organs. The forces that shape the embryo. (Prentice-Hall, Englewood Cliffs) 237 pp. (1969). Trinkaus, J.P.: On the mechanism of metazoan cell movements. In: The Cell Surface in Animal Embryogenesis and Development, eds. G. Poste and G.L. Nicolson (Elsevier/North-Holland Publishing Company, Amsterdam) pp. 225-329 (1976). Trinkaus, J.P.: Protrusive activity of the cell surface and initiation of cell movement during morphogenesis. Exp. Biol. Med. 10, 130-173 (1985). Vakaet, L.: Some new data concerning the formation of the definitive endoblast in the chick embryo. J. Embryol. Exp. Morphol. 10, 38-57 (1962). Vakaet, L.: Cinephotomicrographic investigations of gastrulation in the chick blastoderm. Arch. Biol. (Li6ge) 81,387-426 (1970). Vakaet, L.: The initiation of gastrular ingression in the chick blastoderm. Am. Zool. 24. 555-562 (1984a). Vakaet, L.: Early development of birds. In: Chimeras in Developmental Biology, eds. N. Le Douarin and A. McLaren (Academic Press, London) pp. 71-88 (1984b). Vakaet, L.: Morphogenetic movements and fate maps in the avian blastoderm. In: Molecular Determinants of Animal Form. UCLA Symposia on Molecular and Cellular Biology, Vol. 31, ed. G.M. Edelman (Alan R. Liss, New York) pp. 99-110 (1985). Vakaet, L. and E. Hertoghs-De Maere: Jonctions intercellulaires dans le feuillet inf6rieur du jeune blastoderme de Poulet, 6tudi~es au microscope 61ectronique h balayage. C.R. Soc. Biol. (Paris) 167, 1300-1301 (1973). Vakaet, L. and Ch. Vanroelen: Localization of microfilament bundles in the upper layer of the primitive-streak-stage chick blastoderm J. Embryol. Exp. Morphol. 67, 59-70 (1982). Vakaet, L., Ch. Vanroelen and L. Andries: An embryological model of nonmalignant invasion or ingression. In: Cell Movement and Neoplasia, eds. M. De Brabander, M. Mareel and L. De Ridder (Pergamon Press. Oxford) pp. 65-75 (1980). Van Hoof, J. and F. Harrisson: Interaction between epithelial basement membrane and migrating mesoblast cells in the avian blastoderm. Differentiation 32, 120-124 (1986). Van Hoof, J., F. Harrisson, L. Andries and L. Vakaet: Micro- 105 injection of glycosaminoglyean-degrading enzymes in the chicken blastoderm. An ultrastructurai study. Differentiation 31, 14-19 (1986). Van Hoof, J., F. Harrison, L. Andries and L. Vakaet: Microinjection in the chick blastoderm. An improved method to study the extracellular matrix in the living organism. Exp. Cell. Res. 155,278-282 (1984). Vanroelen, Ch. and L.C.A. Vakaet: Incorporation of 3SSsulphate in chick blastoderms during elongation and during shortening of the primitive streak. Roux's Arch. Dev. Biol. 190, 233-236 (1981). Vanroelen, Ch., L. Vakaet and L. Andries: Localization and characterization of acid mucopolysaccharides in the early chick blastoderm. J. Embryol. Exp. Morphol. 56, 169-178 (1980a). Vanroelen, Ch., L. Vakaet and L. Andries: Distribution and turnover of testicular hyaluronidase sensitive macromolecules in the primitive streak stage chick blastoderm as revealed by autoradiography. Anat. Embryol. 159, 361-367 (1980b). Vanroelen, Ch., L. Vakaet and L. Andries: Alcian blue staining during the formation of mesoblast in the primitive streak stage chick blastoderm. Anat. Embryol. 160, 361-367 (1980c). Vanroelen, Ch., P. Verplanken and L.C.A. Vakaet: The effects of partial hypoblast removal on the cell morphology of the epiblast in the chick blastoderm. J. Embryol. Exp. Morphol. 70, 189-196 (1982). Wakely, J. and R.A. Badley: Organization of actin filaments in early chick embryo ectoderm: an ultrastructural and immunocytochemical study. J. Embryol. Exp. Morphol. 69, 169-182 (1982). Wakely, J. and M.A. England: Scanning electron microscopy (SEM) of the chick embryo primitive streak. Differentiation 7, 181-186 (1977). Wakely, J. and M.A. England: Development of the chick embryo endoderm studied by S.E.M. Anat. Embryol. 153, 167-178 (1978). Wakely, J. and M.A. England: Scanning electron microscopical and histological study of the basement membranes in the early chick embryos, Proc. R. Soc. Lond. B 206, 329-352 (1979). Weinberger, C. and I. Brick: Primary hypoblast development in the chick. I. Scanning electron microscopy of normal development. Roux's Arch. Dev. Biol. 191, 119-126 (1982). Weinberger, C. and 1. Brick: Primary hypoblast development in the chick. II. The role of cell division. Roux's Arch. Dev. Biol. 191,127-133 (1982). Weinberger, C., P.L. Penner and I. Brick: Polyingression, an important morphogenetic movement in chick gastrulation. Am. Zool. 24, 545-55 (1984). Weiss, L. and S. Nir: On the mechanism of transfilter induction of kidney tubules. J. Theor. Biol. 78, 11-20 (1979). Wolpert, L.: Positional information and morphogenetic signals: An introduction. In: Cell Interactions in Differentiation, eds. M. Karkinen-J~iaskelainen, L. SaxOn and L. Weiss (Acadenuc Press, London) pp. 79-87 (1977).