Fibronectin-rich fibrillar extracellular matrix controls cell migration during amphibian gastrulation

We have reviewed the evidence supporting the notion that the fibrillar extracellular matrix on the basal surface of the blastocoel roof in amphibian embryos directs and guides mesodermal cell migration during gastrulation. Based on extensive experimental evidence in several different systems, we conclude the following: (i) the fibrillar extracellular matrix contains fibronectin (FN) and laminin. (ii) The fibrils are oriented in such a way as to promote directional migration of mesodermal cells during migration. (iii) We have used several different probes to disrupt the interaction between migrating mesodermal cells and the fibrillar extracellular matrix. These probes include: (a) nucleocytoplasmic and interspecific hybridization. Such embryos have defects in FN synthesis and gastrulation. (b) Fab' fragments of anti-FN and anti-integrin VLA-5 IgGs prohibit mesodermal cell adhesion both in vitro and in vivo and gastrulation is arrested. (c) Peptides containing the RGDS sequence specifically inhibit interactions between migrating mesodermal cells and the FN-fibrillar matrix. (d) Tenascin blocks cell adhesion to FN in vitro and gastrulation in vivo. (e) Antibodies against the cytoplasmic domain of beta 1 integrin, when injected into blastomeres, prevent FN-fibrillogenesis in progeny of injected blastomeres and delay mesodermal cell migration selectively in the progeny of injected blastomeres but not in the uninjected blastomere progeny.     

Exogenous tenascin inhibits mesodermal cell migration during amphibian gastrulation.

We have used amphibian gastrulation as a model system to study the action of the extracellular matrix (ECM) glycoprotein tenascin on mesodermal cell migration. Tenascin function was assayed in vitro during spreading of isolated cells from the dorsal marginal zone (DMZ) and during cell migration from DMZ explants. Plastic coated with bovine fibronectin or gastrula ECM was used as a substratum. In both cases, tenascin added to the medium inhibited spreading and migration of mesodermal cells. In addition, a substratum coated with a mixture of fibronectin and tenascin was found to prevent mesodermal cell migration. Tenascin was also microinjected into the blastocoel cavity of living embryos at the late blastula stage. This led to a complete arrest of gastrulation in more than 80% of the cases. Scanning electron microscopy of fractures from arrested gastrulae showed that mesodermal cell migration was blocked. Similar injection experiments carried out at the middle gastrula stage demonstrated that tenascin is able to inhibit cell migration after cells have already contacted the ECM. Mesodermal cell migration in the presence of tenascin could be restored in vitro and in vivo by the monoclonal antibody mAb Tn68 which is known to mask a cell binding site of the molecule. Finally, tenascin microinjected into the blastocoel of blastula or gastrula stage embryos bound within 15 min to the ECM fibrils at all the stages studied. Our results show that exogenous tenascin can be incorporated into embryonic ECM and interferes in vivo with the interactions of cells with a fibronectin-rich matrix.     

Formation of the Neural Plate and the Mesoderm in Normally Developing Embryos of Xenopus laevis

Normally developing embryos of Xenopus were fixed at various stages between the blastula and early tail bud stage, and their serial sections were examined. The marginal belt of the blastula was characterized by abundance of cells with RNA-rich peripheral cytoplasm called mesoplasm. At the early gastrula stage, the marginal belt was folded into two layers giving rise to mesodermal material and marginal ectoderm. During gastrulation, the mesodermal material, which consisted of RNA-rich cells, spread to enclose the blastocoel and the endoderm, and a large part of it was shifted to the dorsal side of the embryo. It gradually established the mesodermal layer. The notochord was formed on the dorsal lip of the blastopore by involution, separately from preformed mesodermal material. The RNA-rich cells in the marginal ectoderm became columnar, forming a broad belt in the marginal zone. This belt was deformed and shifted to the dorsal side during gastrulation, eventually establishing the neural plate showing quantitative differentiation along the head-tail axis. Possible mechanisms involved in the formation of the neural plate and mesoderm were discussed with reference to the organizer and the mesoplasm.     

The 140-kDa fibronectin receptor complex is required for mesodermal cell adhesion during gastrulation in the amphibian Pleurodeles waltlii

We have studied the localization and function of a 140-kDa glycoprotein complex implicated in cell adhesion to fibronectin- and laminin-rich extracellular matrices in Pleurodeles waltlii gastrulae. In particular, we have shown that antibodies directed against highly purified avian fibronectin (FN) receptor complex cross-react with two major polypeptides of apparent molecular weights of 140,000 and 100,000 and a third minor component of 90,000. Using sections of embryos or whole mounts, we have also discovered that the putative FN receptor is widely distributed on the early embryonic cell surface. We have also found that the basal surface of the roof of the blastocoel, a region particularly enriched in an extracellular matrix consisting of fibronectin- and laminin-rich fibrils, is rich in receptor complex. We have prepared monovalent Fab' fragments of this antibody and have found that they cause detachment of cells previously attached to substrata coated with fibronectin, and they also arrest gastrulation when injected into the blastocoel of early gastrulae. Thus, it appears that the fibronectin receptor complex plays a significant functional role in cell attachment of gastrula-stage cells in vitro and in cell migration in vivo during gastrulation.     

Regulation of cell polarity, radial intercalation and epiboly in Xenopus: novel roles for integrin and fibronectin

Fibronectin (FN) is reported to be important for early morphogenetic movements in a variety of vertebrate embryos, but the cellular basis for this requirement is unclear. We have used confocal and digital time-lapse microscopy to analyze cell behaviors in Xenopus gastrulae injected with monoclonal antibodies directed against the central cell-binding domain of fibronectin. Among the defects observed is a disruption of fibronectin matrix assembly, resulting in a failure of radial intercalation movements, which are required for blastocoel roof thinning and epiboly. We identified two phases of FN-dependent cellular rearrangements in the blastocoel roof. The first involves maintenance of early roof thinning in the animal cap, and the second is required for the initiation of radial intercalation movements in the marginal zone. A novel explant system was used to establish that radial intercalation in the blastocoel roof requires integrin-dependent contact of deep cells with fibronectin. Deep cell adhesion to fibronectin is sufficient to initiate intercalation behavior in cell layers some distance from the substrate. Expression of a dominant-negative beta1 integrin construct in embryos results in localized depletion of the fibronectin matrix and thickening of the blastocoel roof. Lack of fibronectin fibrils in vivo is correlated with blastocoel roof thickening and a loss of deep cell polarity. The integrin-dependent binding of deep cells to fibronectin is sufficient to drive membrane localization of Dishevelled-GFP, suggesting that a convergence of integrin and Wnt signaling pathways acts to regulate radial intercalation in Xenopus embryos.     

Experimental analysis of the extension of the dorsal marginal zone in Pleurodeles waltl gastrulae

The capacity for extension of the dorsal marginal zone (DMZ) in Pleurodeles waltl gastrulae was studied by scanning electron microscopy and grafting experiments. At the onset of gastrulation, the cells of the animal pole (AP) undergo important changes in shape and form a single layer. As gastrulation proceeds, the arrangement of cells also changes in the noninvoluted DMZ: radial intercalation leads to a single layer of cells. Grafting experiments involving either AP or DMZ explants were performed using a cell lineage tracer. When rotated 90 degrees or 180 degrees, grafted DMZ explants were able to involute normally and there was extension according to the animal-vegetal axis of the host. In contrast, neither single nor bilayered explants from AP involutes completely, and neither extends when grafted in place of the DMZ. Furthermore, when inside of the host, these AP grafts curl up and inhibit the closure of the blastopore. Once transplanted to the AP region, the DMZ showed no obvious autonomous extension. DMZs cultured in vitro showed little extension and this only from the late gastrula stage onward. Removal of blastocoel roof blocked involution to a varied extent, depending on the developmental stage of the embryos. From these results, it is argued that differences could well exist in the mechanism of gastrulation between anuran and urodele embryos. That migrating mesodermal cells play a major role in urodele gastrulation is discussed.     

Mesodermal cell migration during Xenopus gastrulation

The adhesive glycoprotein fibronectin (FN), which is a component of the network of extracellular matrix fibrils on the inner surface of the blastocoel roof (BCR), has been proposed to play a major role in directing mesodermal cell migration during amphibian gastrulation. In the first part of this paper, the adhesion of Xenopus mesodermal cells to FN in vitro is examined. Cells from several mesoderm regions, which differ in developmental fate and morphogenetic activity, are able to bind specifically to the RGD cell-binding site of FN. Dorsal mesodermal cell adhesion to FN varies along the anterior-posterior (a-p) axis: adhesion is strongest in the anterior head mesoderm, and gradually decreases posteriorly. This a-p gradient of mesodermal adhesiveness to FN does not change during mesodermal involution, and is reflected in the morphology of mesodermal explants on FN. An a-p strip of mesoderm develops a spreading, leading anterior margin and a compact, retracting posterior end, thus moving slowly and directionally over the FN substrate at about 0.8 micron/min. Although dissociated cells from all levels of the dorsal mesodermal axis adhere to FN, only the anterior, leading prospective head mesoderm cells migrate as single cells on a FN substrate in vitro. Locomotion by means of lamelliform protrusions occurs at an average rate of about 1.5 micron/min. Cells of the more posterior axial mesoderm merely shift position at random without substantial net translocation, and preinvolution mesoderm cells are completely stationary. On the BCR, the in vivo substrate for mesodermal cell migration, dissociated prospective head mesoderm cells spread and migrate as on FN in vitro, at 2.2 microns/min. In the presence of an RGD peptide which inhibits cell-FN interaction, cells remain globular and do not spread. They are still motile, but change direction frequently, which leads to less efficient net translocation. Apparently, interaction with the RGD cell-binding site of FN and concomitant spreading of head mesoderm cells is required for the stabilization of cell locomotion. In contrast to the directional migration of the mesoderm cell population toward the animal pole in the embryo, the pathways of dissociated cells on the BCR are randomly oriented. Coherent explants of migratory mesoderm do not move at all on the BCR, although they translocate on FN in vitro.(ABSTRACT TRUNCATED AT 400 WORDS)     

Directional mesoderm cell migration in the Xenopus gastrula.

The movement of the dorsal mesoderm across the blastocoel roof of the Xenopus gastrula is examined. We show that different parts of the mesoderm which can be distinguished by their morphogenetic behavior in the embryo are all able to migrate independently on the inner surface of the blastocoel roof. The direction of mesoderm cell migration is determined by guidance cues in the extracellular matrix of the blastocoel roof and by an intrinsic tissue polarity of the mesoderm. The mesodermal polarity shows the same orientation as the external guidance cues and is strongly expressed in the more posterior mesoderm. The guidance cues of the extracellular matrix are recognized by all parts of the dorsal mesoderm and even by nonmesodermal cells from other regions of the embryo. The extracellular matrix consists of a network of fibronectin-containing fibrils. The adhesiveness of this matrix does not vary along the axis of mesoderm movement, excluding haptotaxis as a guidance mechanism in this system. However, an intact fibronectin fibril structure is necessary for directional mesoderm cell migration. When the assembly of fibronectin into fibrils is inhibited, mesoderm explants still migrate on the amorphous extracellular matrix, but no longer directionally. It is proposed that polarized extracellular matrix fibrils may normally guide the migrating mesoderm to its target region.     

Mechanisms of mesendoderm internalization in the Xenopus gastrula: lessons from the ventral side.

Two main processes are involved in driving ventral mesendoderm internalization in the Xenopus gastrula. First, vegetal rotation, an active movement of the vegetal cell mass, initiates gastrulation by rolling the peripheral blastocoel floor against the blastocoel roof. In this way, the leading edge of the internalized mesendoderm is established, that remains separated from the blastocoel roof by Brachet's cleft. Second, in a process of active involution, blastopore lip cells translocate on arc-like trails around the tip of Brachet's cleft. Hereby the lower, Xbra-negative part of the lip moves toward the interior, to contribute mainly to endoderm. In contrast, the upper, Xbra-expressing part moves toward the blastocoel roof-apposed surface of the involuted mesoderm, and eventually becomes inserted into this surface. Vegetal rotation and active mesoderm surface insertion persist over much of gastrulation ventrally. Both processes are also active dorsally. In fact, internalization processes generally spread from dorsal to ventral, though at different rates, which suggests that they are independently controlled. Ventrally and laterally, mesoderm occurs not only in the marginal zone, but also in the adjacent blastocoel roof. Such blastocoel roof mesoderm shares properties with the remaining, ectodermal roof, that are related to its function as substratum for mesendoderm migration. It repels involuted mesoderm, thus contributing to separation of cell layers, and it assembles a fibronectin matrix. These properties change as the blastocoel roof mesoderm moves into the blastopore lip during gastrulation.     

Laminin Fibrils in Newt Gastrulae Visualized by the Immunofluorescent Staining

An anastomosing network of extracellular fibrils on the inner surface of the ectoderm layer of amphibian gastrulae has been shown to provide an adequate substratum for attachment and migration by the mesodermal cells. These fibrils contain fibronectin as shown by immunostaining at the light and electron microscope levels. Now we report the presence of laminin, another cell adhesion glycoprotein, as a fibrillar network on the inner surface of the ectoderm layer in gastrulae of the Japanese newt ( Cynops pyrrhogaster ), but its absence on the blastula ectoderm layer, by the immunofluorescent staining using an antiserum specific for mouse laminin. The same antiserum was shown to stain basement membranes of adult newt organs as expected.     

Mediolateral cell intercalation in the dorsal, axial mesoderm of Xenopus laevis

The pattern of mediolateral cell intercalation in mesodermal tissues during gastrulation and neurulation of Xenopus laevis was determined by tracing cells labeled with fluorescein dextran amine (FDA). Patches of the involuting marginal zone (IMZ) of early gastrula stage embryos, labeled by injection of FDA at the one-cell stage, were grafted to the corresponding regions of unlabeled host embryos. The host embryos were fixed at several stages, serially sectioned, and examined with fluorescence microscopy and three-dimensional reconstruction. Patterns of mixing of labeled and unlabeled cells show that mediolateral cell intercalation occurs in the posterior, dorsal mesoderm as this region undergoes convergent extension and differentiates into somites and notochord. In contrast, it does not occur in any dorsoventral sector of the anterior, leading edge of the mesodermal mantle. These results, taken with other evidence, suggest that the mesoderm of Xenopus consists of two subpopulations, each with a characteristic morphogenetic movement, cell behavior, and tissue fate. The migrating mesoderm (1) does not show convergent extension; (2) migrates and spreads on the blastocoel roof; (3) is dependent on this substratum for its morphogenesis; (4) shows little mediolateral intercalation; (5) consists of the anterior, early-involuting region of the mesodermal mantle; and (6) differentiates into head, heart, blood island, and lateral body wall mesoderm. The extending mesoderm (1) shows convergent extension; (2) is independent of the blastocoel roof in its morphogenesis; (3) shows extensive mediolateral intercalation; (4) consists of the posterior, late-involuting parts of the mesodermal mantle; and (5) differentiates into somite and notochord.     

The Dorsalization of Spermann's Organizer Takes Place during Gastrulation in Xenopus laevis Embryos

Suramin, a polyanionic compound, which is thought to inhibit the binding of growth factors to their receptors, prevents the differentiation of the dorsal blastopore lip of early gastrulae into dorsal mesodermal structures as notochord and somites. Suramin treated blastopore lips form ventral mesodermal structures, mainly heart structures. Several cases showed rythmic contractions ("beating hearts"). Of special interest is the fact that blastopore lips isolated from middle gastrulae followed by suramin treatment differentiate in about 50% of the cases brain structures without the presence of notochord. These data suggest that suramin prevents the differentiation of the dorsal blastopore lip into notochord up to the early middle gastrula stage but no longer the formation of head mesoderm, which is the prequisite for the induction of archencephalic brain structures. Treated chordamesoderm with overlaying ectoderm from late gastrulae will differentiate as untreated controls, namely into dorsal axial structures like notochord, somites and brain structures. The results indicate that primarily a more general or ventral mesodermal signal is transferred from the dorsal vegetal blastomeres (Nieuwkoop center) to the dorsal marginal zone. The dorsalization, which enables the blastopore lip to differentiate into head mesoderm and notochord and in turn to acquire neuralizing activity, takes place during the early steps of gastrulation.     

Cell Locomotion and Contact Guidance in Amphibian Gastrulation

Presumptive mesodermal cells in amphibian gastrulae migratefrom the blastopore toward the animal pole by using the innersurface of the ectodermal layer as their substratum. Duringmigration, the mesodermal cells form lamellipodia and filopodiapredominantly in a direction toward the animal pole. There isa network of the extracellular fibrils on the inner surfaceof the ectodermal layer. The fibrils seem to serve as an adequatesubstratum for attachment of the filopodia and locomotion ofthe mesodermal cells. A significant alignment of the fibrilnetwork along the blastopore—animal pole axis suggestsa hypothesis that it directs morphogenetic cell movements bycontact guidance in combination with contact inhibition of movement.New culture conditions allow the gastrula mesodermal cells tomove actively in vitro with a similar cell shape and at a similarrate as in vivo. Such culture conditions enabled an in vitroexperiment to test the hypothesis of contact guidance. Explantedectodermal layers deposit the fibril network on the surfaceof a cover slip. Dissociated gastrula mesodermal cells seededon such a conditioned surface attach to the surface and moveabout actively. A computer analysis of the time—lapsefilms shows that the cell trails are significantly aligned alongthe blastopore—animal pole axis of the ectodermal layerthat conditioned the surface. The deposited fibril network showsthe alignment along the same axis. There is also a tendencyof the mesodermal cells to move in a polarized fashion preferentiallytoward the animal pole. These results support the hypothesisof contact guidance of mesodermal cell migration in vivo byoriented extracellular fibrils     

Xenopus embryonic cell adhesion to fibronectin: position-specific activation of RGD/synergy site-dependent migratory behavior at gastrulation

During Xenopus laevis gastrulation, the basic body plan of the embryo is generated by movement of the marginal zone cells of the blastula into the blastocoel cavity. This morphogenetic process involves cell adhesion to the extracellular matrix protein fibronectin (FN). Regions of FN required for the attachment and migration of involuting marginal zone (IMZ) cells were analyzed in vitro using FN fusion protein substrates. IMZ cell attachment to FN is mediated by the Arg-Gly-Asp (RGD) sequence located in the type III-10 repeat and by the Pro-Pro-Arg- Arg-Ala-Arg (PPRRAR) sequence in the type III-13 repeat of the Hep II domain. IMZ cells spread and migrate persistently on fusion proteins containing both the RGD and synergy site sequence Pro-Pro-Ser-Arg-Asn (PPSRN) located in the type III-9 repeat. Cell recognition of the synergy site is positionally regulated in the early embryo. During gastrulation, IMZ cells will spread and migrate on FN whereas presumptive pre-involuting mesoderm, vegetal pole endoderm, and animal cap ectoderm will not. However, animal cap ectoderm cells acquire the ability to spread and migrate on the RGD/synergy region when treated with the mesoderm inducing factor activin-A. These data suggest that mesoderm induction activates the position-specific recognition of the synergy site of FN in vivo. Moreover, we demonstrate the functional importance of this site using a monoclonal antibody that blocks synergy region-dependent cell spreading and migration on FN. Normal IMZ movement is perturbed when this antibody is injected into the blastocoel cavity indicating that IMZ cell interaction with the synergy region is required for normal gastrulation.     

A maternal-effect mutation disturbs extracellular matrix organization in the early Pleurodeles waltl embryo

Summary In the early development of the urodele amphibian Pleurodeles waltl, a fibronectin-containing extracellular matrix underlies the inner face of the blastocoel roof. When gastrulation occurs, the fibronectin fibrils provide a suitable substrate for mesodermal-cell migration. Delay in morphogenetic movements of gastrulation has been described in embryos from mutant females (ac/ac) of Pleurodeles waltl. Studies of abnormal mutant gastrulae with fluorescent lectins and immunostaining for fibronectin reveal that they lack a normal matrix. The fibronectin-containing extracellular material always gives rise to a granular pattern without fibronectin-fibril formation. Fibronectin and ₅₁ syntheses occur normally in maternal-effect embryos. In vitro, mesodermal cells from early mutant gastrulae adhere and migrate on fibronectin-conditioned substrata.     

Mesoderm formation in Eleutherodactylus coqui: body patterning in a frog with a large egg.

The direct developing frog, Eleutherodactylus coqui, develops from a large egg (diameter 3.5 mm). To investigate the effect of egg size on germ-layer formation, we studied mesoderm formation in E. coqui and compared it to that of Xenopus laevis (diameter 1.3 mm). First, we identified the position of prospective mesoderm in the 16-cell E. coqui embryo by cell-lineage tracing. Although the animal blastomeres are small, they form most of the blastocoel roof and make extensive contributions to some mesodermal tissues. Second, we performed recombinant analysis with X. laevis animal caps to define the distribution of mesoderm-inducing activity. Mesoderm-inducing activity in E. coqui was restricted around the marginal zone with strong activity in the superficial cells. Neither the vegetal pole nor the blastocoel floor had activity, although these same regions from X. laevis induced mesoderm. Third, we cloned Ecbra, a homologue of Xbra, an early mesoderm marker in X. laevis. Ecbra was expressed in the marginal ring close to the surface, similar to X. laevis, but E. coqui had weaker expression on the dorsal side. Our results suggest that mesoderm formation is shifted more animally and superficially in E. coqui compared to X. laevis.     

The Spemann organizer meets the anterior‐most neuroectoderm at the equator of early gastrulae in amphibian species

The dorsal blastopore lip (known as the Spemann organizer) is important for making the body plan in amphibian gastrulation. The organizer is believed to involute inward and migrate animally to make physical contact with the prospective head neuroectoderm at the blastocoel roof of mid‐ to late‐gastrula. However, we found that this physical contact was already established at the equatorial region of very early gastrula in a wide variety of amphibian species. Here we propose a unified model of amphibian gastrulation movement. In the model, the organizer is present at the blastocoel roof of blastulae, moves vegetally to locate at the region that lies from the blastocoel floor to the dorsal lip at the onset of gastrulation. The organizer located at the blastocoel floor contributes to the anterior axial mesoderm including the prechordal plate, and the organizer at the dorsal lip ends up as the posterior axial mesoderm. During the early step of gastrulation, the anterior organizer moves to establish the physical contact with the prospective neuroectoderm through the “subduction and zippering” movements. Subduction makes a trench between the anterior organizer and the prospective neuroectoderm, and the tissues face each other via the trench. Zippering movement, with forming Brachet's cleft, gradually closes the gap to establish the contact between them. The contact is completed at the equator of early gastrulae and it continues throughout the gastrulation. After the contact is established, the dorsal axis is formed posteriorly, but not anteriorly. The model also implies the possibility of constructing a common model of gastrulation among chordate species.     

Establishment of substratum polarity in the blastocoel roof of the Xenopus embryo

The fibronectin fibril matrix on the blastocoel roof of the Xenopus gastrula contains guidance cues that determine the direction of mesoderm cell migration. The underlying guidance-related polarity of the blastocoel roof is established in the late blastula under the influence of an instructive signal from the vegetal half of the embryo, in particular from the mesoderm. Formation of an oriented substratum depends on functional activin and FGF signaling pathways in the blastocoel roof. Besides being involved in tissue polarization, activin and FGF also affect fibronectin matrix assembly. Activin treatment of the blastocoel roof inhibits fibril formation, whereas FGF modulates the structure of the fibril network. The presence of intact fibronectin fibrils is permissive for directional mesoderm migration on the blastocoel roof extracellular matrix.     

Guidance of mesoderm cell migration in the Xenopus gastrula requires PDGF signaling

In vertebrates, PDGFA and its receptor, PDGFRalpha, are expressed in the early embryo. Impairing their function causes an array of developmental defects, but the underlying target processes that are directly controlled by these factors are not well known. We show that in the Xenopus gastrula, PDGFA/PDGFRalpha signaling is required for the directional migration of mesodermal cells on the extracellular matrix of the blastocoel roof. Blocking PDGFRalpha function in the mesoderm does not inhibit migration per se, but results in movement that is randomized and no longer directed towards the animal pole. Likewise, compromising PDGFA function in the blastocoel roof substratum abolishes directionality of movement. Overexpression of wild-type PDGFA, or inhibition of PDGFA both lead to randomized migration, disorientation of polarized mesodermal cells, decreased movement towards the animal pole, and reduced head formation and axis elongation. This is consistent with an instructive role for PDGFA in the guidance of mesoderm migration.     

Cell adhesion to extracellular matrix in normal Rana pipiens gastrulae and in arrested hybrid gastrulae Rana pipiens female X Rana esculenta male

Rana pipiens eggs fertilized by Rana esculenta sperm (ESC) hybrid embryos develop until gastrulation in control Rana pipiens embryos (PIP) and then show morphogenetic arrest. After arrest, ESC do not gastrulate but live for 5 days as blastula-like embryos. We studied the distribution of fibronectin (FN)-containing fibrils and integrin (INT) in PIP and ESC. There are many FN-fibrils in PIP organized in anastomosing networks radiating away from the center of individual cells and across intercellular boundaries. ESC have fewer fibrils compared to PIP. These fibrils are first located between cells in disorganized arrays. After arrest in ESC, when PIP are Stage 14 neurulae, many more FN-fibrils appear. INT-staining occurs in both embryos in similar patterns. In xenoplastic transplantations, we found that the extracellular matrix on the inner surface of the ESC blastocoel roof serves as a substratum for PIP cell migration. In an in vitro assay, we found more cell adhesion to FN-substrata in PIP than in ESC. Cell locomotion rates on FN-substrata were 1.70 +/- 0.85 microns/min for PIP but only 0.46 +/- 0.56 microns/min for ESC. We also found that the inner surface of the blastocoel roof from ESC can not promote cell adhesion and locomotion when Stage 11 fragments are used for conditioning but that Stage 14 fragments can deposit a FN-fibril-rich extracellular matrix which supports PIP mesodermal cell migration at a rate of 1.26 +/- 0.38 microns/min.