Defective nephrin trafficking caused by missense mutations in the NPHS1 gene: insight into the mechanisms of congenital nephrotic syndrome

Congenital nephrotic syndrome of the Finnish type (CNF or NPHS1) is an autosomal recessive kidney disorder resulting in severe proteinurea and renal dysfunction. Although the disease occurs predominantly in the Finnish population, many cases in other populations have also been reported. The disease gene (NPHS1) encodes nephrin, a podocyte transmembrane protein that is an essential component of the podocyte slit diaphragm, the renal ultrafilter. Since the discovery of the gene, many mutations have been reported in the NPHS1 gene in patients with diverse ethnic background. A surprisingly large number of these mutations are missense mutations resulting in single amino acid substitutions. In order to study the pathomechanism of these missense mutations, we have investigated the fate of 21 such mutations hitherto identified in NPHS1 patients. Immunostaining of stable transfected cells expressing the nephrin mutants demonstrated that most of the mutants showed only endoplasmic reticulum (ER) staining and no detectable cell surface localization. Immunoelectron microscopy of cells expressing the wild-type and a mutant nephrin further confirmed that the mutant nephrin could be abundantly found in the ER but not on the plasma membrane. Subcellular fractionation of wild-type and a mutant cell line clearly showed an altered subcellular distribution and molecular mobility of the mutant nephrin. In summary, our data indicate that a defective intracellular nephrin transport, most likely due to misfolding, is the most common consequence of missense mutations in NPHS1.

Molecular basis of glomerular permselectivity

Recent discoveries in kidney research have given new insights into the molecular make-up of the glomerular filter and mechanisms of permselectivity. The identification of mutations in the genes for glomerular basement membrane type IV collagen has thus demonstrated the central role of the glomerular basement membrane as the structural skeleton of the glomerular capillary. Regional deterioration of this framework not only leads to proteinuria, but also to significant leakage of red blood cells into the urinary space. Tracer studies and the characterization of other glomerular basement membrane components, such as proteoglycans, have also emphasized the role of the glomerular basement membrane in the permselectivity process. However, more recent studies on nephrin, a key component of the slit diaphragm, as well as the podocyte and slit diaphragm-associated intracellular proteins, CD2-associated protein, podocin and alpha-actinin-4, have emphasized the role of the slit diaphragm as a central size-selective filtration barrier. These data have provided a completely new understanding of the mechanisms of proteinuria, both in inherited and acquired diseases. In this review, we present the recent progress made in the characterization of proteins that are important for glomerular permselectivity.

The murine nephrin gene is specifically expressed in kidney, brain and pancreas: inactivation of the gene leads to massive proteinuria and neonatal death

A mouse model for congenital nephrotic syndrome (NPHS1) was generated by inactivating the nephrin gene (Nphs1) in embryonic stem cells by homologous recombination. The targeting construct contained the Escherichia coli lacZ gene as a reporter for the Nphs1 promoter. Mice homozygous for inactivated Nphs1 were born at an expected frequency of 25%. Although seemingly normal at birth, they immediately developed massive proteinuria and edema and died within 24 h. The kidneys of null mice exhibited enlarged Bowman's spaces, dilated tubuli, effacement of podocyte foot processes and absence of the slit diaphragm, essentially as found in human NPHS1 patients. In addition to expression in glomerular podocytes, the reporter gene was expressed in the brain and pancreas of (+/-) and (-/-) mice. In the brain, expression was localized to the ventricular zone of the fourth ventricle, the developing spinal cord, cerebellum, hippocampus and olfactory bulb. In the cerebellum, the expression was seen in radial glial cells. Neither anatomical nor morphological abnormalities were observed in the brains of null mice.

Role of nephrin in cell junction formation in human nephrogenesis

Nephrin is a cell adhesion protein located at the slit diaphragm area of glomerular podocytes. Mutations in nephrin-coding gene (NPHS1) cause congenital nephrotic syndrome (NPHS1). We studied the developmental expression of nephrin, ZO-1 and P-cadherin in normal fetal kidneys and in NPHS1 kidneys. We used in situ hybridization and immunohistochemistry at light and electron microscopic levels. Nephrin and zonula occludens-1 (ZO-1) were first expressed in late S-shaped bodies. During capillary loop stage, nephrin and ZO-1 localized at the basal margin and in the cell-cell adhesion sites between developing podocytes, especially in junctions with ladder-like structures. In mature glomeruli, nephrin and ZO-1 concentrated at the slit diaphragm area. P-cadherin was first detected in ureteric buds, tubules, and vesicle stage glomeruli. Later, P-cadherin was seen at the basal margin of developing podocytes. Fetal NPHS1 kidneys with Fin-major/Fin-major genotype did not express nephrin, whereas the expression of ZO-1 and P-cadherin was comparable to that of control kidneys. Although early junctional complexes proved structurally normal, junctions with ladder-like structures and slit diaphragms were completely missing. The results indicate that nephrin is dispensable for early development of podocyte junctional complexes. However, nephrin appears to be essential for formation of junctions with ladder-like structures and slit diaphragms.

Congenital nephrotic syndrome (NPHS1): features resulting from different mutations in Finnish patients

BACKGROUND

Congenital nephrotic syndrome (NPHS1) is a rare disease inherited as an autosomally recessive trait. The NPHS1 gene mutated in NPHS1 children has recently been identified. The gene codes for nephrin, a cell-surface protein of podocytes. Two mutations, named Fin-major and Fin-minor, have been found in over 90% of the Finnish patients. In this study, we correlated the NPHS1 gene mutations to the clinical features and renal findings in 46 Finnish NPHS1 children.

METHODS

Clinical data were collected from patient files, and kidney histology and electron microscopy samples were re-evaluated. The expression of nephrin was studied using immunohistochemistry, Western blotting, and in situ hybridization.

RESULTS

Nephrotic syndrome was detected in most patients within days after birth regardless of the genotype detected. No difference could be found in neonatal, renal, cardiac, or neurological features in patients with different mutations. Nephrin was not expressed in kidneys with Fin-major or Fin-minor mutations, while another slit diaphragm-associated protein, ZO-1, stained normally. In electron microscopy, podocyte fusion and podocyte filtration slits of various sizes were detected. The slit diaphragms, however, were missing. In contrast to this, a nephrotic infant with Fin-major/R743C genotype expressed nephrin in kidney had normal slit diaphragms and responded to therapy with an angiotensin-converting enzyme inhibitor and indomethacin.

CONCLUSIONS

The most common NPHS1 gene mutations, Fin-major and Fin-minor, both lead to an absence of nephrin and podocyte slit diaphragms, as well as a clinically severe form of NPHS1, the Finnish type of congenital nephrotic syndrome.

Discovery of the congenital nephrotic syndrome gene discloses the structure of the mysterious molecular sieve of the kidney

The molecular nature of the glomerular slit diaphragm, the site of renal ultrafiltration, has until recently remained a mystery. However, the identification of the gene affected in congenital nephrotic syndrome has revealed the presence of a novel protein, possibly specific for the slit diaphragm. This protein, which has been termed nephrin, is a transmembrane protein that probably forms the main building block of an isoporous zipper-like slit diaphragm filter structure. Defects in nephrin lead to abnormal or absent slit diaphragm leading to massive proteinuria and renal failure. The discovery of nephrin sheds new light on the glomerular filtration barrier, provides new insight into the pathomechanisms of proteinuria, and even opens up possibilities for the development of novel therapies for this common and severe kidney complication.

Developmental regulation and ultrastructure of glycogen deposits during murine tooth morphogenesis

The distribution and ultrastructure of glycogen deposits were investigated in the murine tooth germ by histochemical periodic acid-Schiff (PAS) staining and transmission electron microscopy. Lower and upper first molars were examined in mouse embryos at embryonic days 11.5-17 (E11.5-E17) and in 2-day-old postnatal (P2) mice. The oral and dental epithelia and the mesenchymal cells were generally PAS-positive during tooth morphogenesis. PAS-negative cells were present at E13 in the distal tip of the tooth bud epithelium and in the contacting mesenchyme, and this complete lack of PAS reactivity continued in the dental papilla mesenchyme and inner enamel epithelium during the cap and bell stages. The lack of glycogen deposits in the interacting epithelium and mesenchyme during early morphogenesis may be associated with their demonstrated high signaling activities. Mesenchymal cells in the dental follicle consistently possessed small clusters or large pools of glycogen, which disappeared by P2. Since an intense PAS reaction was seen in mesenchymal cells at future bone sites, the glycogen in the dental follicle cells may be associated with their development into hard-tissue-forming cells. Ultrastructural observation of the enamel organ cells from the cap to early bell stages (E14-E15) revealed the occurrence of glycogen pools, which were associated with the Golgi apparatus and with vesicles having amorphous contents. Glycogen particles were also occasionally present inside vesicles or in the extracellular matrix. These may be associated with the exocytosis of glycosaminoglycan components into extracellular spaces and the formation of the stellate reticulum.

Nephrin is specifically located at the slit diaphragm of glomerular podocytes

We describe here the size and location of nephrin, the first protein to be identified at the glomerular podocyte slit diaphragm. In Western blots, nephrin antibodies generated against the two terminal extracellular Ig domains of recombinant human nephrin recognized a 180-kDa protein in lysates of human glomeruli and a 150-kDa protein in transfected COS-7 cell lysates. In immunofluorescence, antibodies to this transmembrane protein revealed reactivity in the glomerular basement membrane region, whereas the podocyte cell bodies remained negative. In immunogold-stained thin sections, nephrin label was found at the slit between podocyte foot processes. The congenital nephrotic syndrome of the Finnish type (NPHS1), a disease in which the nephrin gene is mutated, is characterized by massive proteinuria already in utero and lack of slit diaphragm and foot processes. These features, together with the now demonstrated localization of nephrin to the slit diaphragm area, suggests an essential role for this protein in the normal glomerular filtration barrier. A zipper-like model for nephrin assembly in the slit diaphragm is discussed, based on the present and previous data.

Ultrastructural localization of beta-actin and amphoterin mRNA in cultured cells: application of tyramide signal amplification and comparison of detection methods

We describe a nonradioactive preembedding in situ hybridization protocol using digoxigenin-labeled RNA probes and tyramide signal amplification to increase the sensitivity of detection. The protocol is sensitive enough for electron microscopic localization of endogenous messenger RNAs encoding beta-actin and amphoterin. Three visualization methods were compared: diaminobenzidine enhanced by nickel, Nanogold enhanced by silver and gold toning, and fluorescently labeled tyramides. Diaminobenzidine and Nanogold can be used in both light and electron microscopy. The nickel-enhanced diaminobenzidine was the most sensitive visualization method. It is easy to accomplish but a drawback is poor spatial resolution, which restricts its use at high magnifications. Nanogold visualization has considerably better spatial resolution and is therefore recommended for electron microscopy. Fluorescent tyramides, especially TRITC-tyramide, offer a good detection method for fluorescence and confocal microscopy. The methods were used to localize amphoterin and beta-actin mRNAs in motile cells. Both mRNAs were found in the soma and cell processes. In double labeling experiments, beta-actin mRNA localized to filamentous structures that also contained ribosomal proteins. Especially in the cortical cytoplasm, beta-actin mRNA was associated with actin filaments. Direct localization to microtubules was only rarely seen. (J Histochem Cytochem 47:99-112, 1999)

Glial-cell-line-derived neurotrophic factor is required for bud initiation from ureteric epithelium

The shapes of different organs can be explained largely by two fundamental characteristics of their epithelial rudiments - the pattern of branching and the rate of proliferation. Glial-cell-line-derived neurotrophic factor (GDNF) has recently been implicated in the development of metanephric ureteric epithelium (Pichel, J. G., Shen, L., Sheng, H. Z., Granholm, A.-C., Drago, J., Grinberg, A., Lee, E. J., Huang, S. P., Saarma, M., Hoffer, B.J., Sariola, H. and Westphal, H. (1996). Nature 382, 73–76; Sanchez, M.P., Silos-Santiago, I., Frisen, J., He, B., Lira, S.A. and Barbacid, M. (1996). Nature 382, 70–73; Vega, Q.C., Worby, C.A., Lechner, M.S., Dixon, J.E. and Dressler, G.R. (1996). Proc. Nat. Acad. Sci. USA 93, 10657–10661). We have analysed the target cells of GDNF and the manner in which it controls ureteric development, and have compared it with other growth factors that have been associated with the regulation of branching morphogenesis, namely hepatocyte growth factor (HGF) and transforming growth factor-beta1 (TGFbeta1). We show that GDNF binds directly to the tips of ureteric bud branches, and that it has the ability to promote primary ureteric buds from various segments of Wolffian duct and to attract ureteric branches towards the source of GDNF. It increases cell adhesion, but is not obviously mitogenic for ureteric cells. The data indicate that GDNF is required primarily for bud initiation. Comparison of GDNF, HGF and TGFbeta1 suggests that the latter act later than GDNF, and may represent a partially redundant set of mesenchyme-derived growth factors that control ureteric development. Thus, GDNF is the first defined inducer in the embryonic metanephric kidney.

Morphological and immunohistochemical characteristics of axial structures in the transitory human tail

Ultrastructural relationships between the notochord and neighboring spinal cord were examined during the regression of the human tail. Also, the presence of certain extracellular matrix components in the notochord was immuno-histochemically analysed in the 4th to 12th week old embryos. At the early stages, a close apposition of the notochord to the spinal cord exists in the entire tail region. The external surface of both structures is covered with a continuous basal lamina. The narrow tissue interspace contains interdigitating cell processes and both amorphous and fibrillar extracellular matrix material. With advancing embryonic age, separation of the two structures occurs in craniocaudal direction and the widening interspace becomes occupied by mesenchymal cells. During tail regression and spinal cord retraction, the appearance of large intercellular spaces and cell degeneration takes place in both tissues. With age, the extracellular matrix of the notochord, predominantly the perinotochordal sheath, increases in amount and antigenic complexity. While the intensity of laminin, collagen type IV and type III expression rises continuously during the period examined, the expression of fibronectin begins first at later stages, after the separation of the notochord from the spinal cord. The possible developmental significance of the described phenomena in the regression of the posterior end of the human tail remains to be elucidated.

Contractile proteins and nonerythroid spectrin in oogenesis of Xenopus laevis

The distribution of contractile proteins, actin and myosin, and an actin-binding protein, spectrin, was studied in oogenesis of Xenopus laevis. These proteins are present in oocytes already at the previtellogenic stages, which are characterized by their diffuse distribution. The localization of proteins changed with the beginning of vitellogenesis. At all vitellogenic stages, including the fully grown oocyte, animal-vegetal differences were noted in localization of actin and myosin: in the animal hemisphere they appear as fibrillar-like structures, while in the vegetal one they are localized around the yolk platelets. By the end of the oocyte's growth, a cortical gradient appeared: predominant localization of actin and myosin in the cortical area. As the oocyte maturation proceeded, the distribution of actin and myosin again became diffuse and nonuniform, so that a cortical gradient appears. At the beginning of vitellogenesis spectrin is distributed as a network all over the ooplasm, while in the fully grown oocyte it is localized mostly in the subcortical area of the animal hemisphere and, as individual inclusions, in other regions of the oocyte. No spectrin is found by the end of maturation. Actin, myosin, and spectrin are also present in the oocyte's nuclei. Changes in the distribution of contractile proteins and spectrin during oocyte maturation are discussed with respect to the development of cortical contractility, as well as to the changes in spatial distribution of yolk platelets and regional sensitivity of the maturing oocyte to cytochalasin B.

[A morphological study of the keratin cytoskeleton of the oocyte from the clawed toad using heterologous monoclonal antibodies]

Heterologous monoclonal antibodies E2 (against rat cytokeratin 8) and OSC-1 (against cytoskeletal preparations of mouse oocytes) were used to study the presence and distribution of cytokeratins in Xenopus oocytes during their maturation and growth. To improve visualization of cytokeratins, more adequate methods of oocyte fixation and processing were developed. The results on distribution of cytokeratins obtained with the above mentioned antibodies were compared with one another and with published data. It was found that data obtained by different authors, who used different types of antibodies and different methods for fixation and visualization of Xenopus oocyte cytokeratins, are often contradictory and inconsistent. Along with that, common features of cytokeratin distribution are revealed that have been noticed by every author: animal-vegetal asymmetry of cytokeratin distribution and the existence of two keratin domains, cortical and ooplasmic ones. Specific for the cortical domain of the animal hemisphere is the arrangement of filaments in parallels to the surface, whereas filaments of the ooplasmic domain are oriented radially. The vegetal hemisphere is characterized by the presence of a network formed by filaments with different orientation. We also describe certain peculiarities of cytokeratin distribution in previtellogenic oocytes, as well as disintegration of the keratin system by the end of oocyte maturation.

Spinal cord-notochord relationship in normal human embryos and in a human embryo with double spinal cord

Spinal cord-notochord relationship was analyzed histologically and immunohistochemically in normal human conceptuses between the 4-8 developmental weeks and in a 8-week embryo with double spinal cord. In the early 4-week embryo, the gradual closure of the neural tube along the cranio-caudal body axis was paralleled by the differentiation of the median hindge point cells at the ventral midline of the tube and by its temporary close association with the notochord. During the 5th-8th developmental weeks, the neuroepithelium differentiating into three distinct layers was accompanied by a solid, ventromedially positioned notochord. In the abnormal 8-week embryo, the additional spinal cord was located ventrolaterally from the vertebral column. Both spinal cords appeared bilaterally asymmetric, with their floor and roof plates irregularly formed. An abnormally enhanced pattern of neuroepithelial differentiation characterized their dorsal parts. Furthermore, additional spinal nerves and ganglia and an abnormal bony structure were associated with the spinal cord positioned outside the vertebral column. The underlying vertebral bodies were misshaped and contained scattered supernumerary groups of notochord cells. Our investigation underlines the importance of the notochord-neural tube relationship in the morphogenesis of the spinal cord. We suggest that the double spinal cord was induced by the split notochord.

30-kDa heparin-binding protein of brain (amphoterin) involved in neurite outgrowth. Amino acid sequence and localization in the filopodia of the advancing plasma membrane

A cDNA library constructed from mRNA of rat brain was used to clone the cDNA that encodes the 30-kDa heparin-binding protein (amphoterin) that is developmentally regulated in brain and enhances neurite outgrowth in cerebral neurons. cDNA and peptide sequencing identified a dipolar sequence that has been previously found in studies of high mobility group 1 protein: the 184-amino acid cationic region is followed by a cluster of 30 anionic residues. The mRNA encoding amphoterin is also developmentally regulated; it is strongly reduced in quantity after the rapid perinatal growth phase of the rat brain. Anti-synthetic peptide antibodies raised according to the sequence of amphoterin were shown to bind specifically to the protein isolated from brain, and were used to detect amphoterin in subcellular fractions and in immunostaining of cells. Amphoterin was found in the cytoplasm of the cell soma, in the cell processes, and the substrate-attached material. In cells that are at an active stage of spreading and extending their cytoplasmic processes amphoterin was especially associated with plasma membrane filopodia. The distinct localization to the filopodia of the advancing plasma membrane suggests that endogenous amphoterin has a role in the extension of neurite-type cytoplasmic processes in developing cells. This inference is further supported by the finding that both anti-amphoterin and the anti-synthetic peptide antibodies in the culture media strongly inhibit the outgrowth of cytoplasmic processes.

Basement membrane matrices in mouse embryogenesis, teratocarcinoma differentiation and in neuromuscular maturation

In this paper we discuss studies on basement membrane and interstitial matrix molecules in early development and teratocarcinoma differentiation. In the early embryo a compartmentalization of newly formed cell types takes place immediately by formation of basement membranes. The stage-specific developmental appearance of extracellular matrix molecules such as type IV collagen, laminin, entactin, fibronectin and proteoglycans seems to reflect a diversified role of extracellular matrices already in the earliest stages of development. In teratocarcinoma cultures the appearance and composition of extracellular matrices during the differentiation of endoderm cells closely resembles that found in the early embryo. Also in this respect the teratocarcinoma system can be used as a model for studies on early development. In later developmental phenomena other matrix molecules can also be of importance. Merosin, a novel tissue-specific basement membrane-associated protein that appears during muscle and nerve maturation is an example of such molecules.

Ultrastructure of oral leukoplakia and lichen planus. II. A correlated scanning and transmission electron microscopic study of epithelial surface cells

Nine cases of homogenous leukoplakia and 21 cases of lichen planus (11 reticular and 10 erosive) were studied under the scanning and transmission electron microscopes. Characteristic differences between leukoplakia and lichen planus in surface patterns of the epithelial surface cells were noted. Microridges in leukoplakia form parallel and anastomosing rows whereas the picture in lichen planus is irregular with microridges varying in width. The transmission electron microscopic findings confirm and correlate with these findings, the microridges representing invaginations of the plasma membrane. The keratin patterns exhibited by the surface cells were also characteristic regarding type and site of lesion. Scanning electron microscopy can be used as an additional diagnostic aid in differentiating these lesions.

Ultrastructure of oral leukoplakia and lichen planus. I. Basal region and inflammatory cells

Nine cases of homogenous leukoplakia and 21 cases of lichen planus (11 reticular and 10 erosive) were studied under the electron microscope. The changes found in leukoplakia were limited to occasional breaks in the basal lamina and modest changes in the cytoplasm of the basal cells. The basal lamina in lichen planus was found to exhibit 3 distinctly different pictures reflecting the clinical types of lichen planus. The structural abnormalities in the basal cells increased with increasing severity of the lesions. The inflammatory infiltrate in lichen planus was found to contain mainly small-to-medium-sized lymphocytes but plasma-cells with widely dilated rough endoplasmic reticulum containing a granular substance were frequently seen. Similar changes are seen in other disorders and, therefore, specific diagnostic criteria cannot be established on the basis of the present material.

Basement membrane matrix in vitro: focal binding of exogenous fibronectin to the matrix of teratocarcinoma-derived endodermal cells

Interaction of exogenous fibronectin with the basement membrane-like PYS-2 cell matrix, lacking fibronectin and hyaluronic acid but containing heparan sulfate proteoglycan, was studied in vitro. Both human plasma fibronectin and fibronectin in fetal calf serum bound to PYS-2 matrix; also, fragments of fibronectin containing heparin-binding domains but lacking the collagen-binding domain bound to the matrix. In immunoelectron microscopy the bound fibronectin was found as 20-40 nm globules or patches. Distribution of fibronectin differed from that of laminin and correlated best with that of heparan sulfate proteoglycan. The results suggest that the binding of fibronectin to basement membrane matrices is not due to random adherence but involves specific interactions with other components.

Neuronal differentiation in F9 embryonal carcinoma cells

F9 line embryonal carcinoma cells were induced to differentiate into neural direction by long-term treatment of monolayer cultures with retinoic acid and dibutyryl cyclic AMP. Bi- and multi-polar cells appeared, expressing acetylcholinesterase and neurofilament proteins but not markers of glial differentiation including GFA-protein. Nerve growth factor combined with both retinoic acid and dibutyryl cyclic AMP greatly enhanced the development of neuron-like morphology and induced expression of immunoreactivity to tyrosine hydroxylase as well as to Leu-encephalin-like peptides. Similarly, serotonin-like immunofluorescence but not substance P-like immunoreactivity was demonstrable in such cultures. In addition, synaptic-like vesicles were often found in the processes. Analysis of matrix expression in neuronally differentiated F9 cells revealed marked increase in laminin production, as judged by immunofluorescence and immuno-electron microscopy, but no demonstrable intracellular staining for fibronectin or type IV collagen. The results with neuronal cells contrast with the expression of all the three matrix components in endodermally differentiating F9 cells in the same cultures.

Scanning electron microscopy and X-ray microprobe analysis in detection of acetylcholinesterase in cultured embryonal carcinoma cells

Mouse F9 cells, induced by retinoic acid and dibutyryl cyclic adenosine monophosphate (cAMP) to differentiate into neural-type cells, were incubated for localization of specific acetylcholinesterase (AChE) activity according to the Karnovsky-Roots method where the final enzymatic reaction product is crystalline cupric ferrocyanide and cuprous thiocholine iodide. By scanning electron microscopy (SEM) neural-type cells with long processes were seen. Most of these cells exhibited crystalline precipitates on their surface that in microprobe analysis contained copper, iron, and sulfur. These elements were also detected in some of the neural-type cells that had no visible surface precipitates. Thus, the X-ray analysis also revealed intracellular enzymatic activity. Undifferentiated rounded cells, devoid of AChE activity at the light microscope level, did not show any surface precipitates by SEM and lacked copper, iron, and sulfur emission peaks in the elementary analysis. These results demonstrate that elementary analysis of cytochemical enzymatic reaction products by SEM can be used in identifying cells.

Cell relationships during aggregation between preimplantation embryos and teratocarcinoma-derived cells

Cleavage-stage mouse embryos aggregate and form chimaeric blastocysts with embryonal carcinoma (EC) cells. We used scanning and transmission electron microscopy to study cell relationships during aggregate formation between 8-cell-stage embryos and F9 EC cells. Relations between heterotypic cells were similarly studied in aggregation experiments with embryos and teratocarcinoma-derived visceral (PSA5-E) and parietal (PYS-2) endoderm cells and in experiments with EC cells and endoderm cells.

The embryos and F9 cells always adhered to each other and rapidly formed compacted aggregates. Numerous microvilli and cell processes, originating from both embryo and EC cells, extended between the two cell types during adhesion and early phases of aggregation. The aggregation process involved spreading of the blastomeres on the EC cells. Frequent adherent junctions and close contacts, including possible focal gap or tight junctions were observed between the embryo and F9 cells after 3 h of culture. Apparent gap or tight junctions were infrequent during the early phases of aggregation but during further culture, extensive typical gap junctions were also seen between embryo and EC cells.

The embryos adhered only irregularly and loosely to PSA5-E and PYS-2 cells; this interaction never led to aggregate formation comparable to that seen in the experiments with embryos and EC cells. Close contacts but no gap or tight junctions could be observed between the embryo and endoderm cells. On the other hand, both PSA5-E and PYS-2 cells readily adhered to and aggregated with EC cells.

The present results suggest that microvilli and cell processes mediate membrane interactions during adhesion and early phases of aggregation between embryos and EC cells. During aggregation, blastomeres spread over the EC cells, and rapid formation of adherent junctions and close contacts, including possible focal gap or tight junctions is involved during the early phases of this process. After this initial phase, typical gap junctions are also seen between the embryo and EC cells. Interestingly, adhesive properties of embryo and EC cells differ: the former aggregate only with EC cells, whereas the latter do so also with teratocarcinoma-derived visceral and parietal endoderm cells. Mechanisms operating in the morphogenetic movement of cells in this experimental setup may be involved also in the development of the blastocyst in vivo.

Cell relationships during aggregation between preimplantation embryos and teratocarcinoma-derived cells

Cleavage-stage mouse embryos aggregate and form chimaeric blastocysts with embryonal carcinoma (EC) cells. We used scanning and transmission electron microscopy to study cell relationships during aggregate formation between 8-cell-stage embryos and F9 EC cells. Relations between heterotypic cells were similarly studied in aggregation experiments with embryos and teratocarcinoma-derived visceral (PSA5-E) and parietal (PYS-2) endoderm cells and in experiments with EC cells and endoderm cells. The embryos and F9 cells always adhered to each other and rapidly formed compacted aggregates. Numerous microvilli and cell processes, originating from both embryo and EC cells, extended between the two cell types during adhesion and early phases of aggregation. The aggregation process involved spreading of the blastomeres on the EC cells. Frequent adherent junctions and close contacts, including possible focal gap or tight junctions were observed between the embryo and F9 cells after 3 h of culture. Apparent gap or tight junctions were infrequent during the early phases of aggregation but during further culture, extensive typical gap junctions were also seen between embryo and EC cells. The embryos adhered only irregularly and loosely to PSA5-E and PYS-2 cells; this interaction never led to aggregate formation comparable to that seen in the experiments with embryos and EC cells. Close contacts but no gap or tight junctions could be observed between the embryo and endoderm cells. On the other hand, both PSA5-E and PYS-2 cells readily adhered to and aggregated with EC cells. The present results suggest that microvilli and cell processes mediate membrane interactions during adhesion and early phases of aggregation between embryos and EC cells. During aggregation, blastomeres spread over the EC cells, and rapid formation of adherent junctions and close contacts, including possible focal gap or tight junctions is involved during the early phases of this process. After this initial phase, typical gap junctions are also seen between the embryo and EC cells. Interestingly, adhesive properties of embryo and EC cells differ: the former aggregate only with EC cells, whereas the latter do so also with teratocarcinoma-derived visceral and parietal endoderm cells. Mechanisms operating in the morphogenetic movement of cells in this experimental setup may be involved also in the development of the blastocyst in vivo.

Nerve growth factor induces adrenergic neuronal differentiation in F9 teratocarcinoma cells

Teratocarcinoma cells have been used as a model to study differentiation and development in vertebrates. Treatment with retinoic acid (RA) and dibutyryl cyclic AMP can in some embryonal carcinoma (EC) cell lines lead to neural differentiation, as judged by neurofilament expression and by the induction of enzymes involved in cholinergic transmission. Short-term culture of F9 line cells with RA and dibutyryl cyclic AMP results in a biochemically demonstrable rise in acetylcholinesterase (AChE) activity. We now report that long-term culture of F9 cells with RA and dibutyryl cyclic AMP induces neurofilament expression, demonstrated by immunofluorescence with specific antibodies. Furthermore, if nerve growth factor (NGF) is also added, the developing neurone-like cells exhibit immunoreactivity to tyrosine hydroxylase, a rate-limiting enzyme of catecholamine synthesis specific for adrenergic neurones. Immunoreactivity for Leu-enkephalin-like peptides is also induced. These results suggest that F9 cells can differentiate into cells with adrenergic characteristics.

Steroid metabolism in human teratocarcinoma cell line PA 1

Human ovarian teratocarcinoma cells of line PA 1, (Zeuthen et al., 1979[1]) used as model for early embryonic cells, were analyzed for their in vitro capacity to convert steroids. The cells were incubated for 20 h with radioactive pregnenolone, progesterone, dehydroepiandrosterone, androstenedione, testosterone or estradiol-17 beta, or with non-radioactive progesterone, 6 alpha- or 6 beta-hydroxyprogesterone, 3 beta-hydroxy-5 alpha-pregnan-20-one, dehydroepiandrosterone or estradiol-17 beta. The metabolites were analyzed by thin layer chromatography or studied by gas chromatography-mass spectrometry. The results indicate that PA 1 cells are able to metabolize, although to a restricted amount, a variety of steroids, most markedly progesterone. The metabolites were almost exclusively found in the medium. The main metabolite of progesterone was 3 beta, 6 alpha-dihydroxy-5 alpha-pregnan-20-one. Minor formation of progesterone from pregnenolone could be detected. Human chorionic gonadotropin did not have any effect on pregnenolone metabolism. No formation of estradiol-17 beta or estrone from dehydroepiandrosterone, androstenedione or testosterone could be detected. However, estradiol-17 beta was shown to be converted mainly to estrone. These findings indicate that undifferentiated PA 1 teratocarcinoma cells like certain mouse teratocarcinoma cells, seem not to be steroidogenic but are capable of metabolizing naturally occurring steroid hormones and their precursors.

Identification of hyaluronic acid as a component of human teratocarcinoma-derived cells of line PA 1

Ovarial teratocarcinoma-derived cells (PA 1) of human origin, which are known to be capable of differentiating in vitro, incorporate label from radioactive galactose and 2-amino-2-deoxy-D-glucose into hyaluronic acid. This macromolecule, which is prominent in many differentiating embryonic and malignant cells, was separated from sulfated glycosaminoglycans and from large-sized glycopeptides by chromatography on concanavalin A – Sepharose. It was identified by a number of chemical and enzymatic degradations as well as by several chromatographic methods.

Mannose containing glycopeptides of cells derived from human teratocarcinoma (line PA 1)

Human teratocarcinoma derived cells, line PA 1, maintained in the undifferentiated state, yielded upon exhaustive pronase digestion unusually large glycopeptides (fraction A), which showed on gel filtration an apparent molecular weight larger than 7400. These glycopeptides derived from whole cell proteins carried large-sized oligosaccharides as evidenced by repeated pronase treatments, hydrazinolysis, and beta-elimination experiments. The oligosaccharides consisted of mannose, fucose, galactose, and N-acetylglucosamine. The PA 1 cells contained also oligomannosyl type glycans, presumably linked to asparagine (fraction C glycopeptides). These glycopeptides were strongly bound to Con A--Sepharose and their oligosaccharides were released by endo-beta-N-acetylglucosaminidase H. The liberated glycans ranged from Man5GlcNAc to Man9GlcNAc as analyzed by paper chromatography. "Pulse-chase" experiments suggest that there is a precursor-product relationship between the mannose label in the fraction C (oligomannosyl type) glycopeptides and the fraction A glycopeptides.

Characterization of a human ovarian teratocarcinoma-derived cell line

A cell line (PA I), derived from human ovarian teratocarcinoma cells, was obtained by culturing ascitic fluid cells from a patient with recurrence of malignant ovarian teratoma. During early passages the cultured cells showed a variable morphology, a long doubling time, and a low plating efficiency (2%). After about 50 passages in vitro, a cell population which was more homogeneous and resembled embryonal carcinoma cells were obtained. These cells had a shorter doubling time (26 h), and increased plating efficiency (77%). The early-passage cells were aneuploid (P 24) whereas the late-passage cells had a normal diploid karyotype with one balanced translocation between chromosomes No. 15 and No. 20 (P 224). Details of the karyotype suggest that the cells are heterozygous, i.e. derived from a stage before the first meiotic division. One of the two X chromosomes were inactive, and the cells expressed HLA antigens (A28 and B12), and beta 2-microglobulin. Expression of F9 antigen, characteristic of two-cell and later preimplantation embryos, was absent, while expression of PCC4 antigen, expressed also by blastocysts, was present. This finding suggests that the line might express some embryonic characteristics. The PA I cell line maintained in monolayer cultures showed several characteristics of malignant cells. The proportion of malignant cells increased with successive passages in vitro. The late-passage cells represented a fairly homogenous population of malignant cells similar to embryonal carcinoma cells. Late-passage PA I cells, when seeded under conditions that prevented attachment of cells to the substratum, formed embryoid bodies consisting of an inner core of cells similar to embryonal carcinoma cells, surrounded by a rind of endoderm-like cells. These two cell layers were separated by a basement membrane-like structure containing fibronectin. The core embryonal carcinoma cells expressed high alkaline phosphatase activity whereas the endoderm-like cells had low alkaline phosphatase activity. Embryoid bodies seeded on an adhesive substratum formed polycystic structures divided by layers of epithelial-like cells and containing extracellular fibrils similar to collagen type I or III. In these cultures, further limited differentiation into endoderm-like, epithelial-like cells and pigmented cells was observed. Morphological differenciation of undifferentiated PA I cells into endoderm-like cells in monolayer cultures could be obtained by treatment with BrdUrd or by plating in low serum concentration and at low density. Cells with characteristic fibrillar distribution of fibronectin and actin microfilament bundles were then observed, indicating formation of cells lacking properties of malignant cells. As indicated by these results, the PA I cell line, in spite of a limited capacity to differentiate in vitro, shares some of the properties of mouse teratocarcinoma cell lines and might therefore serve as a useful model for studies on some developmental mechanisms in human cells.

Fibronectin expression is determined by the genotype of the transformed parental cells in heterokaryons between normal and transformed fibroblasts

The expression of fibronectin, a cell surface-associated transformation-sensitive glycoprotein, was studied in hetero- and homokaryons of normal and SV40-transformed human fibroblasts. In immunofluorescence, fibroblast homokaryons had an intense surface-associated and intracelluar fibronectin fluorescence similar to that of normal fibroblasts. Transformed cells and their homokaryons had a minimal surface-associated and a weak intracellular fibronectin fluorescence. In heterokaryons formed between transformed and normal fibroblasts, the expression of fibronectin fell within 24 h to the level of the transformed cell homokaryons. The change was detectable already at 3 h after fusion and was gene-dose dependent. These results show that the transformed genotype determines fibronectin expression in the heterokaryons.

Metabolic inhibitors and kidney tubule induction

The induction of kidney tubules in metanephric mesenchyme has previously been shown to require close contact between the interacting tissues. In our study we show that low concentrations of inhibitors of RNA, DNA and protein synthesis inhibit tubule induction, although they do not seem to prevent the formation of contacts between the interacting tissues. The effective concentrations were about the same as those which inhibited the synthesis of macromolecules. Cycloheximide caused an increased synthesis of RNA. Low concentrations of Mitomycin C inhibited DNA synthesis but not tubule formation. A concentration of the inhibitors which caused weakened induction also caused a marked decrease in leucine incorporation. We concluded therefore, that a decrease in protein synthesis in the inducing tissue is responsible for the inhibition of induction.

Lectin binding in early mouse embryos

Lectin binding in early mouse embryos was used to explore possible changes in cell surface properties during early development. Embryos at different stages of development were stained with fluorescein-coupled Concanavalin A (FITC-Con A) or wheat germ agglutinin (FITC-WGA). In zygote to 8-cell embryos, fixed with paraformaldehyde before staining or stained at 4 degrees C before fixation, the fluorescein-coupled lectins gave a continuous ring-like labelling of cell surfaces. When unfixed embryos stained at 4 degrees C with low concentrations of FITC-Con A (10--20 microgram/ml) were incubated at 37 degrees C they showed a continuous surface labelling at zygote stage but a patch-like aggregation of Con A-binding sites in 2-cell stage and older embryos. Use of higher lectin concentrations ( greater than 100 microgram/ml) did not result in patch formation at 37 degrees C. FITC-WGA gave a spotty but unaggregated labelling pattern also at 37 degrees C. The results show that in the early stages of mouse embryogenesis developmentally associated changes occur in the binding behaviour of Con A but not of WGA.

Fibronectin and the pericellular matrix of normal and transformed adherent cells

Fibronectin is a major glycoprotein component of normal fibroblasts in culture. External fibronectin is predominantly present in a pericellular fibrillar matrix that mediates distant cell-cell and cell-substratum contacts. A small proportion of external fibronectin is closely associated with the plasma membrane. In the matrix, fibronectin is partially disulfide bonded into complexes. Plasma transglutaminase, activated by thrombin, also cross-links external fibronectin into high-molecular-weight covalent complexes. In cultures of normal fibroblasts, pericellular matrix fibronectin displays extensive codistribution with (pro)collagens types I and III. Transformed adherent cells show decreased formation of the fibronectin-collagen matrix. The deficient synthesis of fibronectin and other matrix components and abnormal interactions with the matrix may account for several phenotypic characteristics of transformed cells. The pericellular matrix structure has been prepared by use of deoxycholate and hypotonic medium to solubilize the cells. The matrix contains glycosaminoglycans, procollagens, and fibronectin. The fibronectin codistributes with the procollagens. The matrix may be considered to be an in vitro equivalent of the connective tissue matrix and basal laminae found in vivo. Human sarcoma cells spread rapidly on the prepared matrix and assume an elongated morphology characteristic of normal fibroblasts. The prepared matrix may provide a general tool to study the effects of matrix on cellular behavior and differentiation.

Cross-linking of erythrocyte membrane proteins by periodate and intramembrane particle distribution

Treatment of isolated human erythrocyte membranes at pH 7.4 with 0.1-0.5 mM-sodium periodate specifically cross-linked some of the spectrin polypeptides. Treatment with 2 mM-periodate resulted in complete cross-linking of spectrin and partial cross-linking of other polypeptides. The latter treatment also caused aggregation of the intramembrane particles made visible by freeze-fracturing. When membranes that had been treated with 2 mM-periodate were depleted of spectrin by treatment with 0.1 mM-EDTA, extensive aggregation of the intramembrane particles occurred.

Fluorescent antibodies and lectins stain intracellular structures in fixed cells treated with nonionic detergent

Nonionic detergent (NP40) treatment of paraformaldehyde-fixed normal and SV40-transformed human fibroblasts resulted in intracellular penetration of two chosen fluorescent antibodies and Concanavalin A (Con A). After the detergent treatment nuclear SV40 T antigen, cytoplasmic fibronectin glycoprotein and Con A binding sites could be visualized in fluorescence microscopy. The lowest NP40 concentration which made fixed cells permeable was 0.05%. The morphology of cells was preserved better by this new method than by conventional fixation methods, such as acetone treatment. In scanning electron microscopy the surface of the fixed NP40-treated cells had only small rugosities and fine pores. The subsurface cytoskeleton especially was well preserved and had a more distinct fine structure. The improved morphology made it possible to detect a similar distribution of fibronectin and Con A binding sites in the perinuclear endoplasmic reticulum regions.

External fibronectin of cultured human fibroblasts is predominantly a matrix protein

The distribution of a major glycoprotein (fibronectin) of human fibroblast cultures was studied in immunoelectron microscopy with peroxidase- or ferritin-labeled antibodies. External fibronectin was visualized in pericellular structures, in some areas on the growth substratum, and to a lesser degree in close association with the upper and lower surface membranes of the cell. The pericellular fibronectin-containing structures consisted of amorphous or vaguely fibrillar material forming strands or patches, 50-500 nm in diameter; the structures appeared to mediate distant cell-to-cell and cell-to-substrate contacts. When in close association with the plasma membrane, fibronectin markers were seen as discrete patches. The exact relationship between this form of fibronectin and the plasma membrane, however, remained open. Filamentous material was commonly seen in the cortical cytoplasm under patches of membrane-associated fibronectin. The distribution that we observed is consistent with the proposed roles of fibronectin in cell interactions with neighboring structures and with its presence in vivo as an extracellular glycoprotein in connective tissue matrix and basal laminae.

Lectin-binding sites are found in rat liver cell plasma membrane only on its extracellular surface

In the present study ultrastructural localization of binding sites for 5 lectins was studied in rat liver cell surface membrane fractions. For this purpose ferritin-coupled Concanavalin A, wheat germ agglutinin, soybean agglutinin, Ricinus communis agglutinin 120 and Lotus tetragonolobus agglutinin I were used as probes for mannose, N-acetyl glucosamine, N-acetyl galactosamine, galactose and fucose moieties in glycoproteins and glycolipids. Although recent reports suggest presence of glycogroups on the cytoplasmic surface of cellular membranes ultrastructural identification of membrane surfaces in the present study indicated an asymmetric localization of lectin-binding sites exclusively on the extracellular side of the membranes.

Distribution of glycophorin on the surface of human erythrocyte membranes and its association with intramembrane particles: an immunochemical and freeze-fracture study of normal and En(a-) erythrocytes

Human erythrocyte membranes of the En(a-) blood group lack the major sialoglycoprotein (glycophorin). By absorption of a crude antiglycophorin antiserum with En(a-) membranes a specific antiglycophorin antiserum was obtained. By immune electron microscopy we showed that glycophorin is randomly distributed on the surface of normal erythrocytes. When polycationized ferritin, which mainly binds to glycophorin, was used as a marker a similar even labeling of normal erythrocyte membranes was seen. En(a-) membranes bound much less of this marker. In freeze-fracturing the intramembrane particles of both membrane types had a similar distribution and appeared in equal amounts. However, partial removal of spectrin from these membranes, followed by incubation at pH 6 resulted in more extensive aggregation of the particles in En(a-) membranes than in normal membranes. The results may be interpreted as glycophorin contributing by electrostatic repulsion to the random distribution of the intramembrane particles in normal cells. This repulsion is weakened in in En(a-) cells by the lack of glycophorin.

Changes in the distribution of a major fibroblast protein, fibronectin, during mitosis and interphase

The distribution of a major fibroblast protein, fibronectin, was studied by immunofluorescence and immunoscanning electron microscopy in cultures of human and chicken fibroblasts during different phases of the cell cycle. The main findings were: (a) In interphase cells, the intensity of surface-associated fibronectin fluorescence correlated with that of intracellular fibronectin fluorescence. (b) The intensity of the fluorescence of both surface-associated and intracellular fibronectins was not changed in cells that were synthesizing DNA. (c) Mitotic cells had reduced amounts of surface-associated but not of intracellular fibronectin. The surface fibronectin that remained on meta-, ana-, or telophase cells had a distinct punctate distribution and was also localized to strands attaching the cells to the substratum. Fibronectin strands first reappeared on the surface of flattening cytoplasmic parts of telophase cells. (d) Fibronectin was also detected in extracellular fibrillar material on the growth substratum, particularly around dividing cells. Thus, surface-associated fibrillar fibronectin was present during G(1), S, and G(2) but in cells undergoing mitosis the distribution was altered and the amount appeared to be reduced. The observations on the distribution of surface-associated fibronectin suggest that rather than being involved in growth control this fibronectin plays a structural role in interactions of cells with the environment.