This paper forms part of our study of the extracellular matrix and its role in the morphogenesis of the brain during the period of neurulation in the rat embryo. Using indirect immunofluorescence with polyclonal antibodies, we present here a descriptive study of the distribution of the matrix glycoproteins fibronectin, laminin and entactin.

The observed distribution of the fibronectin matrix implicates it in providing a structural element in several morphologically active sites; in addition our observations support the previously suggested involvement of fibronectin in the migration of neural crest cells. Entactin was present only in the basement membranes in conjunction with laminin which was not itself confined to these regions. Laminin was also identified within the mesenchymal extracellular matrix, and its general distribution confirms the previously documented role of laminin in maintaining epithelial structure and organization. No patterning in the distribution of these three glycoproteins could be correlated with the change in shape of the neural epithelium associated with either tube formation or neuromere morphogenesis.

Extracellular materials play important roles in epithelial morphogenesis and cell migration, as demonstrated for example by studies on salivary gland morphogenesis (Bernfield, Banerjee, Koda & Rapraeger, 1984) and avian neural crest cell migration (Newgreen & Thiery, 1980). We have previously investigated the distribution and morphogenetic functions of hyaluronate and proteoglycans in neurulation, neural crest cell migration and neuromere formation in the cranial region of rat embryos. Hyaluronate is essential for the creation of extracellular spaces around mesenchyme cells, and is particularly localized in the mesenchyme subjacent to the neural epithelium (Solursh & Morriss, 1977; Morriss & Solursh, 1978a,b). Recent results (Morriss-Kay, Tuckett & Solursh, 1986) suggest that its major role is related to the control of mesenchymal cell number rather than cell migration or neuroepithelial morphogenesis. Proteoglycans (particularly chondroitin sulphate proteoglycans) in the neuroepithelial basement membrane play an essential role in the thickening and curvature of this epithelium during neurulation (Morriss-Kay & Crutch, 1982). Glycosaminoglycans do not appear to be involved in the development of the cranial neuromeres (Tuckett, 1984).

We now extend these studies of the extracellular matrix to include fibronectin, laminin and entactin. Fibronectin is a glycoprotein of the extracellular matrix with the ability to form meshworks; neural crest cells preferentially adhere to it in vitro (Greenberg, Seppa & Hewitt, 1981); it promotes cell migration, and binds to collagen, hyaluronate and proteoglycans (Yamada, Hayashi & Akiyama, 1982). Laminin is a glycoprotein of the basal lamina (Timpl et al. 1979). It mediates the binding of epithelial cells to type IV procollagen in the basal lamina and is probably essential for the maintenance of epithelial structure (Terranova, Rohrbach & Martin, 1980). Entactin is a sulphated glycoprotein; it is mainly found in basement membranes but at times also contributes to the extracellular matrix of mesenchymal tissues (Hogan, Taylor, Kurkinen & Couchman, 1982). We describe here the distribution of these three molecules in the cranial region and heart of rat embryos during and immediately after the period of cranial neurulation, as revealed by indirect immunofluorescence. The primary aim was to discover any correlations between the distribution of these substances and morphogenetic events which occur within these regions, at this period of embryogenesis. Previous mammalian studies have been concerned with the stages immediately prior or subsequent to this period (Leivo, Vaheri, Timpl & Wartiovaara, 1980; Wan, Tsung-Chieh, Chung & Damjanov, 1984; Sternberg & Kimber, 1986).

Preparation of specimens

Wistar strain rat embryos were explanted in phosphate-buffered saline (PBS) on either day 9 or day 10 of pregnancy (day of positive vaginal smear = day 0). The extraembryonic membranes were removed from day-10 embryos; day-9 embryos were left intact within their membranes (4-to 5-somite stage). The number of somites present was counted before fixation.

The embryos were fixed in modified St Marie’s fixative (Sainte-Marie, 1962) according to the method described by Icardo & Manasek (1983, 1984); the fixation period was either 1–2 h at 4°C, or overnight at −20°C. The embryos were dehydrated at 4°C in two changes of absolute ethanol followed by two changes of xylene (10 min each change) and allowed to reach room temperature before embedding in paraffin wax.

The embedded embryos were stored at 4°C until they were sectioned. Serial sections of 10 μm thickness were mounted on glass slides and stored at 4 °C until they were required for immunohistochemistry.

Labelling of sections

The sections were hydrated according to the method described by Icardo & Manasek (1983, 1984). The sections were incubated with primary antibody at a 1:50 dilution in PBS. Control sections were incubated with either preimmune serum or PBS instead of the primary antibody (no difference was observed in the background fluorescence between these two solutions). After several washes in PBS, the sections were incubated with the secondary antibody at a 1:50 dilution in PBS; the sections were thoroughly washed in PBS before mounting in u.v.-inert aqueous mountant (Gurr). Both of the incubations were performed in a humid atmosphere, at 38°C, for 30 min. Some sections were double-labelled, i.e. the primary and secondary antibody incubations were repeated using a second set of non-crossreactive antibodies. The primary antibodies were: goat anti-rat fibronectin (CP Laboratories); rabbit anti-mouse laminin (Bethesda Research Laboratories); and rabbit anti-mouse entactin (gift from Brigid Hogan, NIMR, London). The secondary antibodies used were: FITC-conjugated rabbit anti-goat IgG (Miles or Sigma); rhodamine-conjugated rabbit anti-goat IgG (Miles); FITC-conjugated goat anti-rabbit IgG (Miles or Sigma).

The specificity of the two commercially obtained primary antibodies (anti-fibronectin and anti-laminin) was ascertained by preabsorption with fibronectin (Sigma) or laminin (Bethesda Research Laboratories) (Fig. 1); in addition antibody specificity was determined by immunoblotting and the purity of the antigens was ascertained by gel electrophoresis (not illustrated here).

Fig. 1.

Coronal sections along the neural tube of 12-somite-stage embryos illustrating the fluorescence obtained with antisera previously preabsorbed with either fibronectin or laminin. Bar, 50/zm. (A) Anti-laminin preabsorbed with laminin; no non-specific fluorescence. (B) Anti-fibronectin preabsorbed with fibronectin; no non-specific fluorescence. (C) Anti-laminin preabsorbed with fibronectin. FITC-labelled antilaminin fluorescence of the mesenchymal extracellular matrix is punctate whilst within the neuromere basement the fluorescence is continuous. (D) Anti-fibronectin preabsorbed with laminin. FITC-labelled anti-fibronectin fluorescence within the mesenchymal extracellular matrix is generally fibrillar; around the neuromeres the basement membrane fluorescence is continuous.

Fig. 1.

Coronal sections along the neural tube of 12-somite-stage embryos illustrating the fluorescence obtained with antisera previously preabsorbed with either fibronectin or laminin. Bar, 50/zm. (A) Anti-laminin preabsorbed with laminin; no non-specific fluorescence. (B) Anti-fibronectin preabsorbed with fibronectin; no non-specific fluorescence. (C) Anti-laminin preabsorbed with fibronectin. FITC-labelled antilaminin fluorescence of the mesenchymal extracellular matrix is punctate whilst within the neuromere basement the fluorescence is continuous. (D) Anti-fibronectin preabsorbed with laminin. FITC-labelled anti-fibronectin fluorescence within the mesenchymal extracellular matrix is generally fibrillar; around the neuromeres the basement membrane fluorescence is continuous.

The slides were stored in black plastic bags at 4°C. Sections were viewed with an Olympus BH2 microscope fitted with a reflected light fluorescence attachment; Fujichrome ASA 400 positive film was used to photograph the sections for reference, and Ilford HP5 film was used for publication photographs.

Neural epithelium

Throughout the period of cranial neurulation, the neuroepithelial cells did not stain for fibronectin, laminin or entactin. A fibronectin-rich and laminin-rich basement membrane were observed at all stages of development and at all levels within the embryo. Entactin basement membrane staining was very much weaker: at the earliest stage studied (4-somite stage), entactin fluorescence was confined to the forebrain and midbrain regions; however, by the 8-somite stage the fluorescence had progressed more caudally into the hindbrain, although this was very much weaker than the staining seen in the 4-somite stage embryos and was also of a patchy nature. By the 12-somite stage, the entactin basement membrane staining had disappeared at some forebrain and hindbrain levels, the greatest fluorescence (although very weak compared with fibronectin and laminin fluorescence) being observed within the midbrain. In embryos with 15–16 somites there was no epithelial basement membrane staining with entactin antibodies.

With the development of the optic sulci within the expanding forebrain, a regional difference in the staining of fibronectin and the other two glycoproteins was observed. The optic sulci bulge outwards into the cranial mesenchyme and around the bulge, the basement membrane was deficient in both laminin and entactin fluorescence but stained strongly for fibronectin. Basement membrane fluorescence for laminin and entactin remained around the neural groove region and in the region of future forebrain neural fold apposition. This patterning of glycoprotein distribution around the optic sulci persisted throughout the period of cranial neurulation and was not dependent on the plane of sectioning (Fig. 2).

Fig. 2.

Transverse sections through the forebrain/midbrain region of a 10-somite-stage embryo (A,B) and a 12-somite-stage embryo (C,D). Bar, 100 μm. (A) FITC-labelled anti-laminin. Staining is discontinuous around one optic sulcus and is not present around the other. (B) Rhodamine-labelled anti-fibronectin. (C) FITC-labelled anti-laminin. Neuroepithelial basement membrane stains intensely at the lateral margin in the region of the forebrain neuropore and around the neural groove; towards the evaginating optic sulcus this becomes thinned and discontinuous (arrows). (D) Rhodamine-labelled anti-fibronectin. Basement membrane staining remains continuous around the optic sulcus. Laterally the mesenchymal matrix fluorescence is punctate, whilst medially around the notochord (n) the fluorescence is more fibrillar in form.

Fig. 2.

Transverse sections through the forebrain/midbrain region of a 10-somite-stage embryo (A,B) and a 12-somite-stage embryo (C,D). Bar, 100 μm. (A) FITC-labelled anti-laminin. Staining is discontinuous around one optic sulcus and is not present around the other. (B) Rhodamine-labelled anti-fibronectin. (C) FITC-labelled anti-laminin. Neuroepithelial basement membrane stains intensely at the lateral margin in the region of the forebrain neuropore and around the neural groove; towards the evaginating optic sulcus this becomes thinned and discontinuous (arrows). (D) Rhodamine-labelled anti-fibronectin. Basement membrane staining remains continuous around the optic sulcus. Laterally the mesenchymal matrix fluorescence is punctate, whilst medially around the notochord (n) the fluorescence is more fibrillar in form.

Changes in fibronectin staining intensity were also observed at different locations within the neuroepithelial basement membrane. At certain levels within the midbrain and hindbrain, the lateral margins of the neural epithelium were deficient in fibronectin fluorescence and this was associated with the presence of neural crest cells in the immediately adjacent mesenchyme (Fig. 3A,C). After emigration of the neural crest cells to more distal locations, the fibronectin fluorescence was restored at the lateral margins. In several sections anti-laminin antibodies demonstrated a similar deficiency in the basement membrane associated with the emergence and emigration of neural crest cells (Fig. 3B). With the onset of forebrain apposition, the basement membrane of the apposing neural folds was found to fluoresce more strongly for fibronectin than at earlier stages of development and compared with other areas of the forebrain; this increase in fluorescence became even more obvious by the 12-somite stage of development and persisted after apposition had occurred (Fig. 4).

Fig. 3.

Transverse section of a 12-somite-stage embryo cut immediately anterior to the optic pits. At the site of emergence of the neural crest cells, at the lateral margin of the neural epithelium, basement membrane fluorescence is discontinuous. Bar, 50 μm. (A) Rhodamine-labelled anti-fibronectin. (B) FITC-labelled anti-laminin. (C) Rhodamine-labelled anti-fibronectin. High-power view showing the strands of fibronectin fluorescence which remain at the site of crest cell emergence. Neural crest cells (arrowed) are not stained.

Fig. 3.

Transverse section of a 12-somite-stage embryo cut immediately anterior to the optic pits. At the site of emergence of the neural crest cells, at the lateral margin of the neural epithelium, basement membrane fluorescence is discontinuous. Bar, 50 μm. (A) Rhodamine-labelled anti-fibronectin. (B) FITC-labelled anti-laminin. (C) Rhodamine-labelled anti-fibronectin. High-power view showing the strands of fibronectin fluorescence which remain at the site of crest cell emergence. Neural crest cells (arrowed) are not stained.

Fig. 4.

Neural epithelium at the anterior neuropore (10-somite stage). At the lateral-most part of the neuroepithelial basement membrane there is an increase in fibronectin fluorescence. Rhodamine-labelled anti-fibronectin. Bar, 50 μm.

Fig. 4.

Neural epithelium at the anterior neuropore (10-somite stage). At the lateral-most part of the neuroepithelial basement membrane there is an increase in fibronectin fluorescence. Rhodamine-labelled anti-fibronectin. Bar, 50 μm.

In coronal sections stained with fibronectin and laminin antibodies, there was no discontinuity or variation in intensity of fluorescence either within the forebrain, or more caudally surrounding the neuromeres (Fig. 1C,D).

Surface ectoderm

The surface ectodermal cells did not stain with antibodies to the three glycoproteins but there was a glycoprotein-rich basement membrane. Fibronectin was present in the basement membrane at all levels within the embryo and at all stages of development, and the fluorescence was generally strong. Laminin was also present within the basement membrane although there was a variation in staining intensity apparent at the later stages of development: at the level of the pharyngeal arches the fluorescence was patchy and in some sections disappeared altogether. Entactin fluorescence of the basement membrane was very faint, only just above the background level and was only discernible in a very few sections in early-somite-stage embryos. By the 12-somite stage, more entactin was detectable but this was generally confined to the rostral-most levels of the embryo.

The basement membrane fluorescence was continuous with the basement membrane at the lateral margins of the neural folds, except where neural crest cells were emerging from the epithelium; at the level of the optic pits, the basement membranes of surface ectoderm and neural epithelium apparently abutted, resulting in a localized increase in intensity.

Foregut endoderm

With fibronectin and laminin antibodies some of the strongest fluorescence was observed in the basement membrane of the gut endoderm; the endodermal cells themselves did not stain for any of the glycoproteins.The fluorescence for fibronectin and laminin was greatest in the basement membrane of the ventral wall of the foregut. This fluorescence was continuous with the fluorescence associated with the dorsal mesocardium; the basement membrane and dorsal mesocardium were connected by a fibrous reticulum. The intense foregut fluorescence for fibronectin and laminin persisted throughout the period of cranial neurulation. Entactin staining in the foregut basement membrane was very weak and at the 4-somite stage was confined to a few sections. At later stages more of the foregut was entactin-positive, the weak fluorescence being located ventrally, but without a reticular connection with the adjacent dorsal.mesocardium.

Mesenchymal extracellular matrix and somites

Entactin was never observed within the mesenchymal extracellular matrix. The distribution and intensity of stain for laminin and fibronectin varied at different levels of the embryo and at different stages of development.

At the 4-somite stage, there was a punctate staining of the extracellular matrix with both fibronectin and laminin antibodies, which decreased in a rostrocaudal direction. The staining obtained for laminin was much less intense than that for fibronectin and at caudal levels there was no laminin fluorescence. Using fibronectin antibodies it was found that the somatopleuric mesodermal layer was rich in fibronectin whilst the splanchnopleuric layer was deficient in fibronectin, as were the more medial intermediate and paraxial mesoderm. By the 8-somite stage of development, both fibronectin and laminin fluoresced more intensely. At the forebrain level there was a mesh of fibronectin-rich fibres; this was also found at more caudal levels associated with the primary mesenchyme of the midbrain and hindbrain regions; around the notochord the density of the fibres increased to form a reticulum which connected the notochord with the neural groove (Fig. 5B), and in later stages extended laterally to surround the dorsal aortae. There was no fibrillar fluorescence of the neural crest mesenchyme, the punctate staining persisted in this region. Within the first pharyngeal arch the mesenchymal extracellular matrix displayed punctate fluorescence at the 8-somite stage but by the 10-somite stage this was beginning to be infiltrated by strands of fluorescence, indicative of the development of a fibrous meshwork within this region.

Fig. 5.

(A) FITC-labelled anti-laminin. Transverse section at the level of the first somite (12-somite stage). Sclerotome cells have migrated away from the somite leaving the dermamyotome; fluorescence can be seen in the region of the dermatome basement membrane. Bar, 50 μm. (B) FITC-labelled anti-fibronectin. Section through the midbrain (12-somite stage). Fluorescence around the notochord (n), neural groove (g), foregut (f) and dorsal aortae (a) is more fibrillar than in more lateral regions. Bar, 50 μm.

Fig. 5.

(A) FITC-labelled anti-laminin. Transverse section at the level of the first somite (12-somite stage). Sclerotome cells have migrated away from the somite leaving the dermamyotome; fluorescence can be seen in the region of the dermatome basement membrane. Bar, 50 μm. (B) FITC-labelled anti-fibronectin. Section through the midbrain (12-somite stage). Fluorescence around the notochord (n), neural groove (g), foregut (f) and dorsal aortae (a) is more fibrillar than in more lateral regions. Bar, 50 μm.

Antibodies to laminin stained in a punctate manner throughout the period studied. The intensity of fluorescence increased at the later stages examined, although there continued to be a rostrocaudal decrease in the staining intensity.

At somitic levels, the somite cells showed no fluorescence for either fibronectin or laminin but were surrounded by a fibronectin and laminin-rich basement membrane. In the 4-somite-stage embryos, the more lateral unsegmented mesoderm was fibronectin- and laminin-free although a reticulum of fibronectin fluorescence was seen spreading inwards from the overlying ectoderm and underlying endoderm. Associated with the increase in staining at somitic levels at later stages of development, was the appearance of a more widespread fibronectin- and laminin-rich extracellular matrix. There was no fibrillar material spreading inwards from the ectoderm and endoderm basement membranes in the 8-somite-stage embryos or older, although fibronectin fluorescence within the mesenchymal extracellular matrix was fibrillar. In a 12-somite-stage embryo where the first somites were beginning to disperse (i.e. sclerotome cells had migrated), fluorescence with laminin antibodies was noticeable on the developing dermamyotome cells (Fig. 5A), suggesting that a laminin-rich basement membrane was developing between the dermatome and myotome.

Extraembryonic membranes

The extraembryonic membranes were included with the embryos only at the 4-somite stage. The amnion and extraembryonic mesoderm stained strongly for fibronectin, weakly for laminin and were negative for entactin. In contrast, the extraembryonic endoderm fluoresced strongly for both laminin and entactin but did not stain for fibronectin.

Heart

Initially at the 4-somite stage when the heart consists of a simple longitudinal tube, fibronectin was the main glycoprotein to fluoresce although there was some weak laminin fluorescence associated with the dorsal mesocardium. The dorsal mesocardium and the endocardium fluoresced with fibronectin antibodies. There was no fluorescence of the splanchnopleuric mesoderm. With further development of the embryo, the heart tube forms a loop within the pericardial cavity. The endocardium and pericardium stained strongly for fibronectin and laminin, and only weakly for entactin. Even after the breakdown of the dorsal mesocardium, a strong fluorescence to all three glycoproteins persisted in the region between the ventral gut endoderm and the heart tube. Between the splanchnopleuric mesoderm of the pericardial cavity (the myoepicardial mantle) and the endocardium, a cell-free space was traversed by fibrils of fluorescence. The fibrils of fibronectin and laminin were more numerous, longer and fluoresced more strongly in the truncus and rostral end of the bulbus cordis (Fig. 6A,B); entactin antigenicity also developed with time within this reticulum; at the proximal end of the bulbus cordis (right ventricle) and the left ventricle there was only a very narrow cell-free space with a few fibrillar strands traversing it. At the most advanced stage of development studied here (15-to 16-somite stage), generally the fluorescence within the reticulum was weaker although all three glycoproteins were expressed; the fibronectin fibrils fluoresced much less strongly in these later-stage embryos. Within the developing myocardium there was punctate fibronectin fluorescence and fibrils of laminin fluorescence, laminin being the stronger. Within the walls of the developing ventricles, myocardial trabeculation was associated with an initial increase in fluorescence of both laminin and fibronectin,, and the onset of entactin antigenicity within the myocardium. A peak in fibronectin fluorescence associated with myocardial trabeculation was observed around the 12-somite stage (Fig. 6C), being weaker here at the 15-to 16-somite-stage.

Fig. 6.

Transverse sections through the heart of a 12-somite-stage embryo: (A,B) at the level of the truncus and bulbus cordis; (C) at the level of the ventricular trabeculated myocardium, g, foregut; a, dorsal aorta; b, bulbus cordis; t, truncus. Bar, 100 μm. (A) FITC-labelled anti-laminin. (B) Rhodamine-labelled anti-fibronectin. Note the brighter fluorescence within the reticulum compared with (A). (C) FITC-labelled anti-fibronectin. Within the trabeculated myocardium there is an increase in fibronectin fluorescence which reaches a peak in intensity at this stage of development. Compare with (B) which is of a more rostral level and displays much less myocardial fluorescence.

Fig. 6.

Transverse sections through the heart of a 12-somite-stage embryo: (A,B) at the level of the truncus and bulbus cordis; (C) at the level of the ventricular trabeculated myocardium, g, foregut; a, dorsal aorta; b, bulbus cordis; t, truncus. Bar, 100 μm. (A) FITC-labelled anti-laminin. (B) Rhodamine-labelled anti-fibronectin. Note the brighter fluorescence within the reticulum compared with (A). (C) FITC-labelled anti-fibronectin. Within the trabeculated myocardium there is an increase in fibronectin fluorescence which reaches a peak in intensity at this stage of development. Compare with (B) which is of a more rostral level and displays much less myocardial fluorescence.

We have described here the distribution of three glycoproteins of the cell surface and extracellular matrix (fibronectin, laminin and entactin) in rat embryos during the period of cranial neurulation, as revealed by indirect immunofluorescence. The results are summarized in Tables 1–3.

Table 1.

Summary of the general distribution of fibronectin

Summary of the general distribution of fibronectin
Summary of the general distribution of fibronectin
Table 2.

Summary of the general distribution of laminin

Summary of the general distribution of laminin
Summary of the general distribution of laminin
Table 3.

Summary of the general distribution of entactin

Summary of the general distribution of entactin
Summary of the general distribution of entactin

The fixation and embedding protocol was based on the method described by Icardo & Manasek (1983, 1984). Duband & Thiery (1982) found that Sainte-Marie’s fixative and paraffin embedding, both of which were used in this study, decreased the antigenicity of fibronectin. We obtained a very strong fluorescence for both fibronectin and laminin, suggesting that the antigenicity of both was good in our sections. The fluorescence obtained with entactin antibodies was in general weaker, and was confined to basement membranes. This may reflect a masking of antigenic determinants but we do not believe this to be true; evidence for the validity of our results for entactin comes from the very bright fluorescence which was observed in the parietal endoderm. Reichert’s membrane, the thick basement membrane of the parietal endoderm, is known from other studies to contain entactin (for references see Hogan, Barlow & Kurkinen, 1984).

We found laminin to be a component of almost all basement membranes, whereas entactin was not always present. Entactin was never found in the absence of laminin. These observations are consistent with evidence that entactin forms a stable non-covalent complex with laminin (Carlin, Jaffe, Bender & Chang, 1981; Hogan et al. 1982). Fibronectin was present in all basement membranes at all stages.

In association with formation and further development of the optic sulcus, basement membrane immunoreactivity was incomplete. Elsewhere in the neural epithelium all three glycoproteins were present, but at the two optic sulci both laminin and entactin were missing. Basement membrane modification may therefore be implicated in formation and, or, maintenance of the sharp angle of the sulcus. Morphologically, this bend in the neural epithelium differs from neuroepithelial curvature associated with neural tube formation in that the cells are clearly wedge-shaped with very broad basal surfaces. Cytochalasin D had no effect on optic sulcus shape suggesting that this structure is not microfilament-dependent (Morriss-Kay & Tuckett, 1985). Basement membrane modification has also been observed in relation to cleft formation during branching of the mammary and salivary glands (Bernfield, Banerjee, Koda & Rapraeger, 1984): type IV collagen and chondroitin sulphate proteoglycan were lost, while laminin and other components remained. Laminin is considered to be essential for normal epithelial structure and organization (Terranova et al. 1980), whilst type IV collagen may also be essential since it is present in all basement membranes (Timpl & Martin, 1982). It may be that the loss of one or other of them enables bending to occur by modifying the relationship between the basal epithelial surface and the basement membrane.

We were unable to identify any variation in the distribution of fibronectin, laminin or entactin in relation to neuromere formation. Similarly, we have been unable to correlate the distribution of glycosaminoglycans (using alcian blue histochemistry) with neuromeres (Tuckett, 1984). It seems likely by analogy with other systems that a tissue interaction is involved; however, either this occurs at an early stage, determining the neuromeric pattern well in advance of neuromere morphogenesis, or the molecules involved in the extracellular component of the interaction have yet to be identified.

Coinciding with the onset of neural crest cell migration from the midbrain and hindbrain regions, basement membrane fluorescence was lost from the lateral neural epithelium. Once the crest cells had migrated away, basement membrane fluorescence was restored. This sequence of loss and restoration of the basement membrane supports the view that neural crest cell emigration is quantal, as suggested by fibronectin studies in the chick (Newgreen & Thiery, 1980) and by scanning electron microscopy in the rat embryo (Tan & Morriss-Kay, 1985). In the chick embryo, Newgreen & Thiery (1980) observed differences in neural-crest-related fibronectin distribution at different axial levels during crest cell migration: pioneer crest cells at cranial and sacral levels synthesized fibronectin while crest cells from cervical to lumbar axial levels did not. We also observed differences in neural-crest-associated fibronectin at different axial levels within the cranial region: rostral to the somites there was punctate fluorescence between the neural crest cells, suggesting that they secrete fibronectin; there was no pattern of fluorescence which could be correlated with crest cell migration pathways at somitic levels, suggesting that neural crest cells do not secrete fibronectin at more caudal levels.

Of the three matrix glycoproteins we have studied here fibronectin is the one which is generally associated with cell migration during embryonic development. However recent evidence has led to the hypothesis that fibronectin and laminin may play reciprocal roles in controlling cell movement during the development of the peripheral and central nervous systems respectively. There are three lines of evidence to support this hypothesis: (i) fibronectin and not laminin has been identified along neural crest cell migration pathways prior to the differentiation of the peripheral nervous system (reviewed by Le Douarin, 1984); (ii) in vitro it has been demonstrated that both central and peripheral neurones extend on a laminin-bound substratum, whereas only peripheral neurones extend on a fibronectin-bound substratum (Rogers et al. 1983); (iii) in vivo, Leisi (1985) found transient expression of laminin immunoreactivity within the neural epithelium which coincided with periods of neuronal migration; the earliest stage at which laminin was described occurred at a similar stage of development to the one studied here. However the findings of Leisi are in direct conflict with our observations and of others (Bignami, Chi & Dahl, 1984; Wan et al. 1984; Le Douarin, 1984): laminin immunoreactivity was confined to the basement membranes and the mesenchymal extracellular matrix; thus we conclude that during the period of cranial neurulation, laminin has no specific morphogenetic role.

Apposition of the lateral edges of the neural epithelium to form a closed tube was associated with an increase in basement membrane fibronectin fluorescence in the forebrain region but not elsewhere. Closure of the anterior neuropore occurs separately from closure of the midbrain/hindbrain neuropore, and fusion of the apposing epithelia is not an immediate consequence of apposition (Morriss & New, 1979 and unpublished observations). Adhesion of the apposing apical surfaces may set up tensions requiring that the adhering epithelia are strongly bonded to their basement membrane, and through it to the underlying mesenchyme. An increase in the fibronectin content of the basement membrane could bring this about. Alternatively (or in addition) the increase in fibronectin may be related to the increase in cell–cell contacts (Chen et al. 1978; Furcht, Mosher & Wendelschafer-Crabb, 1978) or to changes in cell alignment and position (Yamada, Olden & Pastan, 1978; Zetter, Martin, Birdwell & Gospodarowicz, 1978) which occur during the fusion process.

Around the notochord a fibronectin-rich reticulum stretched upwards to the neural groove basement membrane and laterally towards the cranial mesenchyme or somites. The mammalian notochord is a flimsy structure compared with that of avian and amphibian embryos (Morriss-Kay, 1981); this reticulum may be important in communication between the notochord and adjacent cells, e.g for maintenance of neural groove shape, or it may provide extra structural support. In addition it is suitably placed for playing a role in the guidance of sclerotome cells in their migration towards the notochord from the somites, as previously described in amphibian embryos (Lofberg et al. 1978), and providing a substrate for their proliferation.

In the heart, the distribution of fibronectin was similar to that reported by Icardo & Manasek (1983) in chick embryos. The extracellular matrix (cardiac jelly) of the reticulum has also been shown to be rich in glycosaminoglycans and collagen, whose temporal and spatial pattern of development suggests that they are synthesized and secreted by the endocardium (Markwald & Adams-Smith, 1972; Manasek et al. 1973; Johnson, Manasek, Vinson & Seyer, 1974; Markwald, Fitzharris & Adams-Smith, 1975; Manasek, 1976). Our observations of a generally low level of myocardial but high endocardial and reticular fluorescence suggest that the same is true for fibronectin, laminin and entactin in the mammalian heart. The fibrillar pattern of laminin fluorescence in the myocardium at the latest stages examined probably indicates the onset of development of a laminin-rich basement membrane secreted by and enclosing the myocardial cells. The pattern of fluorescence in the dorsal mesocardium at early stages suggest that all three glycoproteins are involved in maintenance and change of the position of the heart in relation to the foregut until the dorsal mesocardium breaks down.

The results of this study are consistent with previous reports on the distribution of fibronectin suggesting roles in neural crest cell migration and as a supportive element in the reticulum of the heart. In addition they suggest that it has a supportive role where it forms a reticulum around the notochord, and a role in epithelial fusion during closure of the anterior neuropore. The role of entactin is not clear; its appearance and disappearance from different epithelial basement membranes was not obviously related to any consistent morphogenetic change; the only exception to this was the disappearance of both laminin and entactin from the optic sulcus basement membrane. Laminin was present in all other basement membranes, confirming its essential role in epithelial structure and organization.

We wish to thank Martin Barker for technical assistance and Tony Barclay for photographic assistance; Brigid Hogan for the entactin antibody; Penny Thomas for the gel electrophoresis and immunoblotting; and the MRC for financial support.

Bernfield
,
M.
,
Banerjee
,
S.
,
Koda
,
J.
&
Rapraeger
,
A.
(
1984
).
Remodelling of the basement membrane as a mechanism of morphogenetic tissue interaction
.
In The Role of Extracellular Matrix in Development
(ed.
R. L.
Trelstad
), pp.
545
572
.
New York
:
Liss
.
Bignami
,
A.
,
Chi
,
N.
&
Dahl
,
D.
(
1984
).
First appearance of laminin in peripheral nerve, cerebral blood vessels and skeletal muscle of the rat embryo. Immunofluorescence study with laminin and neurofilament antisera
.
Int. J. Dev. Neurosci
.
2
,
367
376
.
Carlin
,
B.
,
Jaffe
,
R.
,
Bender
,
B.
&
Chang
,
A. E.
(
1981
).
Entactin, a novel basal lamina-associated sulfated glycoprotein
.
J. biol. Chem
.
256
,
5209
5214
.
Chen
,
L. B.
,
Murray
,
A.
,
Segal
,
R. A.
,
Bushnell
,
A.
&
Walsh
,
M. L.
(
1978
).
Studies on intercellular LETS glycoprotein matrices
.
Cell
14
,
377
391
.
Duband
,
J. L.
&
Thiery
,
J. P.
(
1982
).
Distribution of fibronectin in the early phase of avian cephalic neural crest cell migration
.
Devi Biol
.
93
,
308
323
.
Furcht
,
L. T.
,
Mosher
,
D. F.
&
Wendelschafer-Crabb
,
L. G.
(
1978
).
Differences in the ultrastructural localisation of fibronectin on normal and transformed cells
.
Ann. N.Y. Acad. Sci
.
312
,
426
431
.
Greenberg
,
J. H.
,
Seppa
,
S.
&
Hewitt
,
A. H.
(
1981
).
Role of collagen and fibronectin in neural crest cell adhesion and migration
.
Devi Biol
.
87
,
259
266
.
Hogan
,
B. L. M.
,
Taylor
,
A.
,
Kurkinen
,
M.
&
Couchman
,
J.
(
1982
).
Synthesis and localisation of two sulphated glycoproteins associated with basement membranes and the extracellular matrix
.
J. Cell Biol
.
95
,
197
204
.
Hogan
,
B. L. M.
,
Barlow
,
D. P.
&
Kurkinen
,
M.
(
1984
).
Reichert’s membrane as a model for studying the biosynthesis and assembly of basement membrane components
.
Ciba Foundation Symp
.
108
,
60
69
.
Icardo
,
J. M.
&
Manasek
,
F. J.
(
1983
).
Fibronectin distribution during early chick embryo heart development
.
Devi Biol
.
95
,
19
30
.
Icardo
,
J. M.
&
Manasek
,
F. J.
(
1984
).
An indirect immunofluorescence study of the distribution of fibronectin during the formation of the cushion tissue mesenchyme in the embryonicheart
.
Devi Biol
.
101
,
336
346
.
Johnson
,
R. C.
,
Manasek
,
F. J.
,
Vinson
,
N. C.
&
Seyer
,
J. M.
(
1974
).
The biochemical and ultrastructural demonstration of collagen during early heart development
.
Devi Biol
.
36
,
252
271
.
Le Douarin
,
N.
(
1984
).
Cell migrations in embryos
.
Cell
38
,
353
360
.
Leisi
,
P.
(
1985
).
Do neurons in the vertebrate CNS migrate on laminin?
EMBOJ
.
4
,
1163
1170
.
Leivo
,
L
,
Vaheri
,
A.
,
Timpl
,
R.
&
Wartiovaara
,
J.
(
1980
).
Appearance and distribution of collagens and laminin in the early mouse embryo
.
Devi Biol
.
76
,
100
114
.
Lofberg
,
J.
,
Ahlfors
,
K.
&
Fallstrom
,
C.
(
1978
).
Extracellular matrix fibrils and migrating cells in amphibian embryos studied by light microscopy in combination with transmission and scanning electron microscopy (TEM and SEM)
.
J. Ultrastruct. Res
.
63
,
106a
.
Manasek
,
F. J.
(
1976
).
Glycoprotein synthesis and tissue interactions during establishment of the functional embryonic chick heart
.
J. molec. Cell Cardiol
.
9
,
389
402
.
Manasek
,
F. J.
,
Reid
,
M.
,
Vinson
,
R.
,
Seyer
,
J. M.
&
Johnson
,
R.
(
1973
).
Glycosaminoglycan synthesis by the early embryonic chick heart
.
Devi Biol
.
35
,
332
348
.
Markwald
,
R. R.
&
Adams-Smith
,
W. N.
(
1972
).
Distribution of mucosubstances in the developing rat heart
.
J. Histochem. Cytochem
.
20
,
896
907
.
Markwald
,
R. R.
,
Fitzharris
,
T. P.
&
Adams-Smith
,
W. N.
(
1975
).
Structural analysis of endocardial cytodifferentiation
.
Devi Biol
.
42
,
160
180
.
Morriss
,
G. M.
&
New
,
D. A. T.
(
1979
).
Effect of oxygen concentration on morphogenesis of cranial neural folds and neural crest in cultured rat embryos
.
J. Embryol. exp. Morph
.
54
,
17
35
.
Morriss
,
G. M.
&
Solursh
,
M.
(
1978a
).
Regional differences in mesenchymal cell morphology and glycosaminoglycans in early neural-fold stage rat embryos
.
J. Embryol. exp. Morph
.
46
,
37
52
.
Morriss
,
G. M.
&
Solursh
,
M.
(
1978b
).
The role of primary mesenchyme in normal and abnormal morphogenesis of mammalian neural folds
.
Zoon
6
,
33
38
.
Morriss-Kay
,
G. M.
(
1981
).
Growth and development of pattern in the cranial neural epithelium of rat embryos during neurulation
.
J. Embryol. exp. Morph
.
65
Suppl.,
225
241
.
Morriss-Kay
,
G. M.
&
Crutch
,
B.
(
1982
).
Culture of rat embryos with -D-xyloside: evidence of a role for proteoglycans in neurulation
.
J. Anat
.
134
,
491
506
.
Morriss-Kay
,
G. M.
&
Tuckett
,
F.
(
1985
).
The role of microfilaments in cranial neurulation in rat embryos: effects of short-term exposure to cytochalasin D
.
J. Embryol. exp. Morph
.
88
,
333
348
.
Morriss-Kay
,
G. M.
,
Tuckett
,
F.
&
Solursh
,
M.
(
1986
).
The effects of Streptomyces hyaluronidase on tissue organization and cell cycle time in rat embryos
.
J. Embryol. exp. Morph. (submitted)
.
Newgreen
,
D. F.
&
Thiery
,
J. P.
(
1980
).
Fibronectin in early avian embryos: synthesis and distribution along the migratory pathways of neural crest cells
.
Cell Tissue Res
.
211
,
269
291
.
Rogers
,
S. L.
,
Letourneau
,
P. C.
,
Palm
,
S. L.
,
McCarthy
,
J.
&
Furcht
,
L. T.
(
1983
).
Neurite extension by peripheral and central nervous system neurons in response to substratum-bound fibronectin and laminin
.
Devi Biol
.
398
,
212
220
.
Sainte-Marie
,
G.
(
1962
).
A. paraffin embedding technique for studies employing immunofluorescence
.
J. Histochem
.
10
,
250
256
.
Solursh
,
M.
&
Morriss
,
G. M.
(
1977
).
Glycosaminoglycan synthesis in rat embryos during the formation of the primary mesenchyme and neural folds
.
Devi Biol
.
57
,
75
86
.
Sternberg
,
J.
&
Kimber
,
S. J.
(
1986
).
Distribution of fibronectin, laminin and entactin in the environment of migrating neural crest cells in early mouse embryos
.
J. Embryol. exp. Morph. (submitted)
.
Tan
,
S. S.
&
Morriss-Kay
,
G. M.
(
1985
).
The development and distribution of the cranial neural crest in the rat embryo
.
Cell Tissue Res
.
240
,
403
416
.
Terranova
,
V. P.
,
Rohrbach
,
D. H.
&
Martin
,
G. R.
(
1980
).
Role of laminin in the attachment of PAM212 (epithelial) cells to basement membrane collagen
.
Cell
22
,
719
726
.
Timpl
,
R.
&
Martin
,
G. R.
(
1982
).
In Immunochemistry of the Extracellular Matrix
, vol.
II
(ed.
H.
Furthmayr
), pp.
119
150
.
Boca Raton
:
CRC Press
.
Timpl
,
R.
,
Rohde
,
H.
,
Robey
,
P. G.
,
Rennard
,
S. L
,
Foidart
,
J. M.
&
Martin
,
G. R.
(
1979
).
Laminin - A glycoprotein from basement membranes
.
J. biol. Chem
.
254
,
9933
9937
.
Tuckett
,
F.
(
1984
).
Pattern and control of mitosis during cranial neurulation in the rat embryo
.
D. Phil Thesis, University of Oxford
.
Wan
,
Y-J.
,
Tsung-Chieh
,
W.
,
Chung
,
A. E.
&
Damjanov
,
I.
(
1984
).
Monoclonal antibodies to laminin reveal the heterogeneity of basement membranes in the developing and adult mouse tissues
.
J. Cell Biol
.
98
,
971
979
.
Yamada
,
K. M.
,
Olden
,
K.
&
Pastan
,
I.
(
1978
).
Transformation-sensitive cell surface protein: isolation, characterization, and role in cellular morphology and adhesion
.
Ann. N.Y. Acad. Sci
.
312
,
256
277
.
Yamada
,
K. M.
,
Hayashi
,
M.
&
Akiyama
,
S. K.
(
1982
).
Structure and function of fibronectin
.
In Extracellular Matrix
(ed.
S.
Hawkes
and
J. L.
Wang
), pp.
25
34
.
Academic Press
.
Zetter
,
B. R.
,
Martin
,
G. R.
,
Birdwell
,
C. R.
&
Gospodarowicz
,
D.
(
1978
).
Role of the high-molecular-weight glycoprotein in cellular morphology, adhesion and differentiation
.
Ann. N.Y. Acad. Sci
.
312
,
299
316
.