Masking of antigenic sites of fibronectin by glycosaminoglycans in ethanol-fixed embryonic tissue

Summary We studied the interaction between glycosaminoglycans (GAGs) and fibronectin in the basement membrane of the epiblast in the chicken blastoderm using testicular-hyaluronidase digestion of GAGs either on fixed tissue sections or in vivo after microinjection of the enzyme preparation prior to immunostaining for fibronectin. In the choice of fixatives, special attention was paid to their preservation of GAGs. The controls included alcian-blue staining of serial sections to test the efficiency of the digestion, and incubations in the presence of protease inhibitors to abolish contaminating proteolytic activity in the commercial hyaluronidase preparations. The results indicate that fixation in solutions which preserve GAGs, i.e. ethanolic solutions or aqueous solutions containing cetylpyridinium chloride, allows the immunocytochemical demonstration of fibronectin in the basement membrane of the epiblast at the level of the endophyllic crescent, but masks this glycoprotein at the epithelial-mesenchymal interface. As shown by both approaches, this masking of immunoreactivity is reversible. Moreover, the in vivo clearance of GAGs before fixation shows that the masking at the epithelial-mesenchymal interface is not an experimental peculiarity due to the use of a particular technique, but is the consequence of an interaction between GAGs and fibronectin in that particular area of the basement membrane that is used by mesoblast cells as a substrate for migration. The observation that fibronectin may be masked by GAGs in ethanol-fixed tissue — a commonly used fixation method—may require the re-evaluation of some negative results mentioned in the literature.     

Expression of different regional patterns of fibronectin immunoreactivity during mesoblast formation in the chick blastoderm

The appearance and distribution of the extracellular material glycoprotein, fibronectin, was investigated in gastrulating chick embryos using affinity-purified anti-human plasma fibronectin antibodies. Preservation of tissue structure and immunoreactivity was carried out by ethanol/acetic acid fixation or by formaldehyde/glutaraldehyde fixation. Using the former fixation method, fibronectin immunoreactivity was detected (1) at the ventral surface of the upper layer or epiblast, mainly anterior and lateral to Hensen's node, in regions where middle-layer or mesoblast cells are not yet present, and (2) sparsely in extracellular spaces of the deep layer. Using the latter fixation method, fibronectin immunoreactivity was, moreover, found at the entire ventral surface of the upper layer, i.e., also at the epithelial-mesenchymal interface, where a basement membrane was previously described. At the light microscope level, we could not detect significant immunoreactivity in the middle layer. Treatment of sections of ethanol-fixed blastoderms with testicular hyaluronidase before immunostaining for fibronectin partially demasked the antigenic sites of this glycoprotein at the epithelial-mesenchymal interface. The present report indicates that the different regional patterns of fibronectin immunoreactivity in the basement membrane of the upper layer are spatially and temporally correlated with migration and positioning of mesoblast cells. These regional patterns are probably due to differences in the composition of fibronectin-associated material such as chondroitin sulfate A and/or C proteoglycans, and/or hyaluronate, before and after mesoblast expansion, rather than to differences in the distribution of fibronectin itself. In this respect, it is noteworthy that the chemical composition of the basement membrane of an epithelium changes as mesenchyme cells migrate over it. The results also favor the idea that fibronectin is a structural component of the whole basement membrane which is used as a substrate for migration of mesenchymal cells.     

Demonstration of the interaction between glycosaminoglycans and fibronectin in the basement membrane of the living chicken embryo

Summary A method utilizing microinjection of glycosaminoglycan-degrading enzymes in the chicken blastoderm prior to embryo culture and immunostaining for fibronectin have been applied to demonstrate an interaction between glycosaminoglycans and fibronectin in the basement membrane of the epiblast. Fixation of tissue in a mixture of formaldehyde and cetylpyridinium chloride allows detection of fibronectin only in those zones of the embryo that are not colonized by mesoblast cells. The epithelial-mesenchymal interface thus remains unstained. After degradation of glycosaminoglycans in the living organism, it is shown that this particular site, in fact, also contains fibronectin that is masked in vivo by, at least, hyaluronate. This interaction between both compounds is, during gastrulation, constantly correlated with mesoblast migration. Since previous studies have shown that the degradation of hyaluronate determines the behaviour of mesoblast cells, it is proposed that remodelling of the interaction between these compounds is necessary for mesoblast migration to occur.     

Nonuniform distribution of fibronectin during avian limb development

Using indirect immunofluorescence we have examined the distribution of the cell surface and extracellular matrix glycoprotein fibronectin at the epithelial-mesenchymal interface and in the mesenchyme of developing chick and duck wing buds. At all stages examined, in both species, staining for fibronectin is greatly enhanced in the basement membrane subjacent to the apical ectodermal ridge (AER), a site of inductive tissue interaction, relative to the epithelial basement membranes in the noninductive dorsal and ventral limb epithelial-mesenchymal interfaces. In stage 23, 25, and 28 chick limb buds, staining for fibronectin is uniform in the least mature distal mesenchyme, retained between more proximal cells undergoing precartilage condensation and lost in those regions undergoing myogenesis, and persistent in all but the most mature cartilage present at the latest stage examined. These results are consistent with a role for fibronectin in AER-induced limb outgrowth, and with a transient role for the glycoprotein in the formation of the skeletal pattern of the limb.     

Expression and adhesive properties of basement membrane proteins in cerebral capillary endothelial cell cultures

Previous studies have indicated the importance of basement membrane components both for cellular differentiation in general and for the barrier properties of cerebral microvascular endothelial cells in particular. Therefore, we have examined the expression of basement membrane proteins in primary capillary endothelial cell cultures from adult porcine brain. By indirect immunofluorescence, we could detect type IV collagen, fibronectin, and laminin both in vivo (basal lamina of cerebral capillaries) and in vitro (primary culture of cerebral capillary endothelial cells). In culture, these proteins were secreted at the subcellular matrix. Moreover, the interaction between basement membrane constituents and cerebral capillary endothelial cells was studied in adhesion assays. Type IV collagen, fibronectin, and laminin proved to be good adhesive substrata for these cells. Although the number of adherent cells did not differ significantly between the individual proteins, spreading on fibronectin was more pronounced than on type IV collagen or laminin. Our results suggest that type IV collagen, fibronectin, and laminin are not only major components of the cerebral microvascular basal lamina, but also assemble into a protein network, which resembles basement membrane, in cerebral capillary endothelial cell cultures.     

Hypoblast cells of chick pre-streak stage embryos invaded basement membrane analogues in vitro: Implications for hypoblast layer formation

In early chick blastodermal morphogenesis, the hypoblast layer is organized beneath the epiblast and induces an axial structure. However, the origin of hypoblast cells and the mechanism of hypoblast layer formation are poorly understood. We hypothesized that the hypoblast layer is formed by an invasive process across the basement membrane of the juxtaposing epiblast, and tested the idea in vitro . Primary and secondary hypoblast cells from embryos at various pre-streak stages were dissociated into single cells and cultured on reconstituted basement membrane gel, laminin gel or fibronectin gel in the culture medium with or without serum for 24–48 h. As a result, we found that after 24 h of serum-supplemented culture, up to 35% of the hypoblast cells dissolved the gel and made holes on it. Similarly, up to 36% of the hypoblast cells showed invasiveness after 48 h in the serum-free culture. Furthermore, it was observed that Koller's sickle cells, which are regarded to be the progenitors of secondary hypoblast cells, penetrated those gels on which they were seeded. The posterior epiblast cells covering Koller's sickle were also invasive. These results suggest that the presumptive primary hypoblast cells that are known to mingle with epiblast cells invade through the basement membrane to form the hypoblast layer. Furthermore, the present results imply that invasion through the basement membrane may be involved in the formation of Koller's sickle, the anlage of secondary hypoblast.     

Autoradiographic evidence for the sliding of the upper layer over the basement membrane in chicken blastoderms during gastrulation.

An upper layer (epiblast) fragment taken laterally from the Anlage fields of neural plate or chordamesoderm of a quail blastoderm, labelled with 3H-glucosamine, was grafted isotopically (in a similar region), isochronically (at the similar stage of development) and isotropically (with the same caudocranial and dorsoventral polarity) in the epiblast of a mesoblast free area of a chicken blastoderm (St 4-5 Vakaet, 1970: full grown primitive streak). On the autoradiographs of the sections through such cultured blastoderms with fully integrated quail grafts, we observed a labelling of the basement membrane laterally and slightly cranially from the labelled graft in its final position. Since only the epiblast and its basement membrane are involved, the pattern of the observed labelling indicates that the grafted and integrated quail epiblast fragment glides in toto over the mediocaudally localized basement membrane, leaving behind a track of radioactivity. Sliding of whole groups of epiblast cells over the basement membrane seems thus to be a normal phenomenon during avian gastrulation.     

Changes in the distribution of type IV collagen,laminin, proteoglycan,and fibronectin during mouse tooth development

The distribution of certain basement membrane (BM) components including type IV collagen, laminin, BM proteoglycan, and fibronectin was studied in developing mouse molar teeth, using antibodies or antisera specific for these substances in indirect immunofluorescence. At the onset of cuspal morphogenesis, type IV collagen, laminin, and BM proteoglycan were found to be present throughout the basement membranes of the tooth. Fibronectin was abundant under the inner enamel epithelium at the region of differentiating odontoblasts and also in the mesenchymal tissues. After the first layer of predentin had been secreted by the odontoblasts at the epithelial-mesenchymal interface, laminin remained in close association with the epithelial cells whereas type IV collagen, BM proteoglycan, and fibronectin were distributed uniformly throughout this area. Later when dentin had been produced and the epithelial cells had differentiated into ameloblasts, basement membrane components disappeared from the cuspal area. These matrix components were not detected in dentin while BM proteoglycan and fibronectin were present in predentin. The observed changes in the collagenous and noncollagenous glycoproteins and the proteoglycan appear to be closely associated with cell differentiation and matrix secretion in the developing tooth.     

Interaction between epithelial basement membrane and migrating mesoblast cells in the avian blastoderm

We investigated the remodeling of glucosamine-containing basement-membrane components in chimaeric avian embryos during gastrulation. Epiblast grafts metabolically labelled with tritiated glucosamine were excised from gastrulating quail embryos and implanted orthotopically into chicken embryos at the same developmental stage. The chimaerae were allowed to develop in culture for 5-7 h before autoradiographic processing. The resulting autoradiographs not only showed the presence of silver grains in the grafted quail tissue and at the level of its basement membrane, but also revealed labelling in the basement-membrane region of the chicken tissue lateral to the graft, i.e. between the mesoblast and epiblast. This last labelling extended as far as at the edge of the area pellucida, i.e. in a region of chicken tissue situated more laterally than the initial position of the graft. No labelling was observed medial, anterior, or posterior to the graft. This observation argues against the interpretation that our results were due to diffusion of labelled compounds within the basement membrane. We also provide evidence to exclude the possibility that quail epiblast cells migrated on their own underlying basement membrane, leaving behind a carpet of labelled material. Taking into account, firstly, the morphogenetic movements that occur during gastrulation, i.e. the movement of epiblast cells towards the primitive streak where they ingress, and the migration of mesoblast cells along the basement membrane towards the periphery of the area pellucida, and secondly, the medial movement of the basement membrane, it is suggested that mesoblast cells picked up labelled compounds in the basement membrane of the graft and left these behind during their lateral migration.(ABSTRACT TRUNCATED AT 250 WORDS)     

Primitive streak-forming cells of the chick invaginate through a basement membrane

Summary Recently fibronectin was shown to appear in the development of the chick for the first time as a thin band on the epiblastic side facing the hypoblast just prior to primitive streak formation. It was thus suggested that fibronectin might be instrumental in the migration of cells that lead to axis formation during primitive streak formation. In the present work we have examined simultaneously for the presence of fibronectin and the specific basement membrane glycoprotein laminin during primitive streak formation using immunofluorescence methods. Laminin was found to be expressed between the epiblast and the hypoblast of stage XIII¹ chick blastoderms. During the immediately following process of streak formation the laminin was found to be continuously detectable throughout the area covered by the hypoblast, but disrupted on the streak area. Fibronectin was found to co-distribute with laminin in stage XIII and in the early primitive streak chick blastoderms. It is concluded that at stage XIII laminin and fibronectin form part of a basement membrane that is partially disrupted during the immediately following process of primitive streak formation in order to allow the migration of the streak-forming epiblastic cells during this morphogenetic process.     

Epithelial and basement membrane responses to chick embryo primitive streak grafts

Chick embryo primitive streak grafts, placed beneath the epiblast of host embryos, tend to result in the formation of either a neural plate in response to anterior streak grafts, or in de-epithelialization in response to posterior grafts. Ultrastructural and immunocytochemical examination shows that both reactions are preceded by basement membrane disruption and early removal of fibronectin therefrom. This disruption does not occur in response to non-streak grafts. It is suggested that the disruption, evoked by primitive streak cells, is a prerequisite first step, allowing direct graft-epiblast cell contact. This contact elicits a specific cytoskeletal reaction determining the epiblast response.     

Ultrastructural distribution of sulfated glycosaminoglycans in epithelial-mesenchymal interface of developing rat tooth germs

We investigated the ultrastructural distribution of sulfated glycosaminoglycans in the epithelial-mesenchymal interface of tooth germs by use of the high-iron diamine thiocarbohydrazide silver proteinate (HID-TCH-SP) staining and enzymatic digestion method. At an early stage in odontoblast differentiation, HID-TCH-SP stain deposits were sparsely distributed in the basement membrane and in the intercellular spaces. Subsequently, as formation of the initial predentin matrix began, HID-TCH-SP stain deposits were densely distributed in the interfibrillar spaces and the basement membrane. Testicular hyaluronidase digested most of those in the progenitor pre-dentin, whereas those in the region of basal lamina resisted enzymatic digestion. Testicular hyaluronidase-resistant HID-TCH-SP stain deposits were susceptible to heparitinase, indicating that the sulfated glycosaminoglycan in the basal lamina is heparan sulfate. Furthermore, the heparan sulfate tended to be regularly arranged at the sites of internal and external lamina densa. However, as progenitor pre-dentin matrix formation proceeded, the numbers of stain deposits temporarily increased and their distribution pattern became irregular, finally tending to disappear with the disruption of basal lamina.     

Matrix assembly,regulation, and survival functions of laminin and its receptors in embryonic stem cell differentiation

Laminin-1 is essential for early embryonic basement membrane assembly and differentiation. Several steps can be distinguished, i.e., the expression of laminin and companion matrix components, their accumulation on the cell surface and assembly into basement membrane between endoderm and inner cell mass, and the ensuing differentiation of epiblast. In this study, we used differentiating embryoid bodies derived from mouse embryonic stem cells null for gamma1-laminin, beta1-integrin and alpha/beta-dystroglycan to dissect the contributions of laminin domains and interacting receptors to this process. We found that (a) laminin enables beta1-integrin-null embryoid bodies to assemble basement membrane and achieve epiblast with beta1-integrin enabling expression of the laminin alpha1 subunit; (b) basement membrane assembly and differentiation require laminin polymerization in conjunction with cell anchorage, the latter critically dependent upon a heparin-binding locus within LG module-4; (c) dystroglycan is not uniquely required for basement membrane assembly or initial differentiation; (d) dystroglycan and integrin cooperate to sustain survival of the epiblast and regulate laminin expression; and (e) laminin, acting via beta1-integrin through LG1-3 and requiring polymerization, can regulate dystroglycan expression.     

Origin and deposition of basement membrane heparan sulfate proteoglycan in the developing intestine

The deposition of intestinal heparan sulfate proteoglycan (HSPG) at the epithelial-mesenchymal interface and its cellular source have been studied by immunocytochemistry at various developmental stages and in rat/chick interspecies hybrid intestines. Polyclonal heparan sulfate antibodies were produced by immunizing rabbits with HSPG purified from the Engelbreth-Holm-Swarm mouse tumor; these antibodies stained rat intestinal basement membranes. A monoclonal antibody (mAb 4C1) produced against lens capsule of 11-d-old chick embryo reacted with embryonic or adult chick basement membranes, but did not stain that of rat tissues. Immunoprecipitation experiments indicated that mAb 4C1 recognized the chicken basement membrane HSPG. Immunofluorescent staining with these antibodies allowed us to demonstrate that distribution of HSPG at the epithelial-mesenchymal interface varied with the stages of intestinal development, suggesting that remodeling of this proteoglycan is essential for regulating cell behavior during morphogenesis. The immunofluorescence pattern obtained with the two species-specific HSPG antibodies in rat/chick epithelial/mesenchymal hybrid intestines developed as grafts (into the coelomic cavity of chick embryos or under the kidney capsule of adult mice) led to the conclusion that HSPG molecules located in the basement membrane of the developing intestine were produced exclusively by the epithelial cells. These data emphasize the notion already gained from previous studies, in which type IV collagen has been shown to be produced by mesenchymal cells (Simon- Assmann, P., F. Bouziges, C. Arnold, K. Haffen, and M. Kedinger. 1988. Development (Camb.). 102:339-347), that epithelial-mesenchymal interactions play an important role in the formation of a complete basement membrane.     

Correlation between the expression of the HNK-1 epitope and cellular invasiveness in prestreak epiblast cells of chick embryos

During avian gastrulation, certain cells present in the epiblast layer ingress through the basement membrane sealing the basal surface of themselves. Previously we reported that chick prestreak epiblast cells show two different behavioral phenotypes upon reconstituted basement membrane and laminin gel in vitro. Half of the dissociated epiblast cells invade the gel substratum after one-day of culture, whereas the others attach to the gel but do not invade. It is expected that such heterogeneity in the behavior of the epiblast cells reflects some mechanism that sorts the cells into those that will ingress into the blastocoelic cavity and those that will remain in the epiblast layer. To test this hypothesis, we dissociated chick prestreak epiblast cells into single cells, cultured them on the laminin gel, and then stained them with anti-HNK-1 antibody. This antibody binds to an epitope present on half of the prestreak epiblast cells which are thought to differentiate into presumptive mesoendodermal cells. We found that 80% of the invasive epiblast cells were HNK-1-positive whereas 77% of the non-invasive cells were HNK-1 negative. In the case of invasive cells, the edges of the proteolytic holes made by the invasive cells were often stained. These results suggest that the cells expressing the HNK-1 carbohydrate chain are preferentially invasive, and this induces selective ingression of the carrier cells for mesoendodermal differentiation in vivo.     

Glycosaminoglycans in porcine lung: an ultrastructural study using Cupromeronic Blue

Glycosaminoglycans (GAGs) are essential components of the extracellular matrix contributing to the mechanical properties of connective tissues as well as to cell recognition and growth regulation. The ultrastructural localization of GAGs in porcine lung was studied by means of the dye Cupromeronic Blue in the presence of 0.3 M MgCl₂ according to Scott's critical electrolyte concentration technique. GAGs were observed in locations described as follows. Pleura: Dermatan sulphate (DS) and chondroitin sulphate (CS) attached in the region of the d-band of collagen fibrils, interconnecting the fibrils; heparan sulphate (HS) at the surface of elastic fibers and in the basement membrane of the mesothelium and blood vessels. Bronchial cartilage: Abundant amounts of GAGs were observed in three zones: pericellular, in the intercellular matrix and at the perichondrial collagen. By enzyme digestion a superficial cartilage layer with predominantly CS could be distinguished from a deep zone with CS and keratan sulphate. The structure of the large aggregating cartilage proteoglycan was confirmed in situ. Airway epithelium: HS at the whole surface of cilia and microvilli and in the basement membrane of the epithelial cells. Alveolar wall: CS/DS at collagen fibrils, HS at the surface of elastic fibers and in the basement membranes of epithelium and endothelium.     

Ultrastructural immunocytochemical localization of fibronectin deposition during corneal endothelial wound repair. Evidence for cytoskeletal involvement

The distribution of the extracellular matrix (ECM) protein, fibronectin (FN), has been examined ultrastructurally in noninjured and injured rat corneal endothelium in vivo and in vitro by immunoperoxidase cytochemistry. In noninjured endothelia, FN was observed within the rough endoplasmic reticulum (RER) cisternae but not along the cell-Descemet's membrane (DM) interface. Twenty-four and 48 h after a circular freeze injury, immunoperoxidase reaction product was detected at the cell-DM interface as well as within cytoplasmic vesicles and intercellular spaces. By 1 and 2 wk post-injury, a line of reaction product could still be demonstrated at the cell-DM interface and evidence for newly deposited basement membrane material was observed in this region. In order to understand whether fibronectin deposition during wound repair was dependent on cytoskeletal influences, organ culture experiments were performed in which the media was supplemented with either 10(-8) M colchicine or 2.5 X 10(-3) M cytochalasin B. Without inhibitors, injured corneas cultured for 24 h had FN deposition at the cell-DM interface similar to the in vivo results. Corneas cultured in the presence of cytochalasin B also showed FN deposition at the cell-DM interface. However, when injured endothelia were cultured in the presence of colchicine, no reaction product was observed at the cell-DM interface, although it could be detected intracellularly within RER. Incubating the tissues in the presence of puromycin abolished all extracellular and intracellular staining. These results indicate that during wound repair, corneal endothelial cells produce fibronectin and deposit it upon Descemet's membrane by a mechanism that may be mediated by microtubules.     

Redistribution of fibronectin and cytoskeletal proteins during the differentiation of rat mammary tumor cells in vitro

The tumorigenic mammary epithelial stem cell line, Rama 25, is capable of synthesizing and secreting fibronectin but incorporates only small amounts of fibronectin into pericellular material localised in regions of cell-cell and cell-substratum contact. Under certain culture conditions, Rama 25 differentiates into a non-tumorigenic myoepithelial-like cell line, Rama 29, which is capable of retaining fibronectin on the cell surface in characteristic fibrillar formation. The redistribution of fibronectin is accompanied by a reorientation of the cytoskeleton from circular bundles in Rama 25 to parallel arrays of filaments in Rama 29. In vivo, fibronectin is found in the basement membrane of the mammary gland and our in vitro studies lead us to suggest that the mammary myoepithelial cell in vivo synthesizes much of the basement membrane fibronectin.     

Association of mesenchyme with attenuated basement membranes during morphogenetic stages of newt limb regeneration

The interface between epithelium and mesenchyme may be involved in inductive interactions which occur during development. This interface within the growth bud, or blastema, of a regenerating limb has been examined to determine whether changes in basement-membrane structures are visible in regions of putative epithelial-mesenchymal inductive interaction. Regenerating forelimbs of adult newts were fixed by perfusion with osmotically balanced aldehydes. Late-bulb to early-digit stage regenerates were collected and processed either for light and transmission electron microscopy or for scanning electron microscopy. Light microscopy confirmed that regions characterized by increased numbers of subepithelial mesenchymal cells were covered by a diffusely stained basement membrane. Transmission electron microscopy of these regions revealed two structural components of the basement membrane. The thin basal lamina was continuous in all regions of all stages examined, but it was attenuated apically in areas of mesenchymal cell accumulation. The thicker underlying reticular lamina was markedly attenuated in these regions near the blastemal apex. Scanning electron microscopy of de-epithelialized blastemas revealed that, apically, the reticular lamina formed only a delicate lacelike network. On the base of the blastema, it formed a dense fibrillar meshwork which was further organized into a geometric pattern in the adjacent stump skin. Cumulatively, these observations suggest that physical contact between epithelial and mesenchymal cells is not essential at these stages, but that regions of putative epithelial-mesenchymal interaction are characterized by a distinctly diminished reticular lamina. Structural changes in basement-membrane components may be related to termination of local inductive events.     

Epithelial-mesenchymal interactions in the production of basement membrane components in the gut

The production and deposition of extracellular matrix proteins and the cellular origin of type-IV collagen have been analysed immunocytochemically in cocultured or transplanted intestinal epithelial-mesenchymal cell associations. In the first experimental model, rat intestinal endodermal cells were cultured on top of confluent mono-layers of rat intestinal or skin fibroblastic cells. Under these conditions, interstitial matrix and basement membrane proteins were deposited within the fibroblastic layer over the whole culture period; interactions between the epithelial cells and the fibroblastic cell population, whatever their organ of origin, were required for the production of the basement membrane. In addition, its formation was progressive as assessed by the shift of a spot-like labelling to a continuous linear pattern at the epithelial-mesenchymal interface, and paralleled epithelial cell differentiation. In the second experimental model, chick-rat epithelial-mesenchymal recombinants developed as intracoelomic grafts were used, and the immunocytochemical detection of a basement membrane protein, type-IV collagen, was performed with species-specific antibodies. The major role of the mesenchyme in the deposition of type-IV collagen is supported by the fact that anti-chick but not anti-mammalian antibodies stained this antigen in chick mesenchyme-rat endoderm recombinants. These observations emphasize the role of tissue interactions in the formation of a basement membrane and show that the mesenchymal compartment is the principal endogenous source of type-IV collagen.