Astrocyte cell lineage. II. Mouse fibrous astrocytes and reactive astrocytes in cultures have vimentin- and GFP-containing intermediate filaments

Expression of vimentin by cultured astroglia and oligodendroglia

The objective of this study was to determine whether vimentin expression by process-bearing astroglia and oligodendroglia cultured from neonatal rat cerebral cortex resembled that in brain where vimentin is common in immature astroglia and a few subpopulations of mature astroglia, but is absent in oligodendroglia. Vimentin expression was detected by immunofluorescence microscopy using a monoclonal antibody (V9) against porcine lens vimentin in combination with either antiserum against the astroglial marker, glial fibrillary acidic protein (GFAP), or with antiserum against the oligodendroglial marker, galactocerebroside (GC). Specificity of the antivimentin antibody was indicated on immunoblots of process-bearing cell proteins separated by two-dimensional polyacrylamide gel electrophoresis. Enrichment of cultures for either GFAP+ astroglia or GC+ oligodendroglia was achieved by supplementation of the culture medium with fetal calf serum at 10% or 0.5%, respectively. Process-bearing cells maintained in 10% serum exhibited heterogeneity in their expression of GFAP and vimentin. Approximately half of the cells were GFAP+/vimentin+ throughout the 2-week culture period examined. GFAP+/vimentin- cells were a minor population at early times (3-4 days) in culture, but accounted for 40% of process-bearing cells after 2 weeks. Cultures maintained in reduced (0.5%) serum and stained for GC and vimentin also exhibited heterogeneity. Both GC+/vimentin+ and GC+/vimentin- cells were observed, with vimentin+ cells composing two-thirds and one-half of the GC+ population after 3 and 6 days, respectively, in reduced serum. The high incidence of vimentin expression by process-bearing astroglia and oligodendroglia suggests that these cultures contain glia in a relatively early stage of development.(ABSTRACT TRUNCATED AT 250 WORDS)

Changes in ganglioside composition and morphological features during the development of cultured astrocytes from rat brain

Changes in ganglioside content over a period of days were examined in astrocytes obtained via cell passage from rat cerebral cortex. Thin-layer chromatography revealed that, in the astrocytes, ganglioside GM1 was absent, the predominant ganglioside being GM3. Also, an increased GD3 content in long-term astrocyte cultures was detected. The morphological features of astrocytes were also studied using immunoperoxidase staining. Astroglial features were characterized by high levels of glial fibrillary acidic protein (GFAP) and vimentin, which are the major intermediate-filament proteins present in astrocytes at an early culture stage. In long-term-cultured (greater than 7 months) astrocytes, vimentin and GFAP were increased in process-bearing cells. Ganglioside GD3 recognized by R24 monoclonal antibody was also expressed in these cells. These results suggest that the increase of ganglioside GD3 in long-term-cultured astrocytes may be related to the appearance of multistellate cells showing strong reactivity against GFAP and vimentin during development over a specified period in culture.

Expression of 48-kilodalton intermediate filament-associated protein in differentiating and in mature astrocytes in various regions of the central nervous system

In this paper we present evidence that the 48-kD intermediate filament-associated protein (IFAP) is expressed relatively late in maturation of astrocytes, after they have acquired the glial fibrillary acidic protein (GFAP). In the astrocytes of white matter in the cerebellum the GFAP is detected at P3, whereas the 48-kD IFAP is detected only at P11. In the periventricular region and the hippocampus the 48-kD IFAP was detected at P6, long after the appearance of GFAP. In adult mice the 48-kD IFAP was observed in GFAP-positive astrocytes in the white matter of cerebellum, spinal cord, brainstem, and corpus callosum as well as in GFAP-positive cells in the grey matter of cerebral cortex and spinal cord. The 48-kD IFAP was not, however, detected in radial glia and their derivatives, in Bergmann glia or in Müller glia. Thus, not all the GFAP-positive astroglia express the 48-kD IFAP. Similarly, 48-kD IFAP was not detected in cells which were GFAP-negative.

An immunohistological study of 66 ependymomas

Sixty-six ependymomas were examined immunohistologically to determine their distribution of glial fibrillary acidic proteins, S-100 protein and vimentin. The neoplasms were subdivided into four groups: (1) ependymomas from the cauda equina, predominantly of the myxopapillary type; (2) benign ependymomas; (3) malignant ependymomas; and (4) ependymoblastomas. Marked differences in antigen reactivity were observed between each group. The intensity of the reaction with the three antibodies was strongest in malignant ependymomas. Ependymomas from the cauda equina showed a patchy distribution of positivity for the three antigens in cells surrounding blood vessels but there was no staining of collagenous septa or the myxoid areas. In ependymoblastomas, the cells of the rosettes were negative for glial fibrillary acidic protein, but there was focal positivity for vimentin and S-100. Other areas showed tumour cells containing moderate amounts of vimentin and small amounts of S-100, and a few bands of filaments positive for glial fibrillary acidic protein. The cytogenetic and biological implications of these findings are discussed.

Simple method for the culture of glial cells from embryonic rat brain: implications for regional heterogeneity and the radial glial lineage

A simple method is described for the production of glial cell cultures from specific regions of the day 13-15 embryonic rat brainstem and midbrain based on differential cell attachment to a relatively nonadhesive substrate, which inhibits the growth of neurons. Regional differences in the ability of specific populations of brainstem and midbrain cells to attach and spread on the substrate suggest that embryonic glial populations may differ in their cell surface properties even when they derive from the same general area of the developing brain. Based on observations of the spatiotemporal distribution of radial-like glial cells and astrocytes with time in vitro, we suggest that this culture system may prove useful for investigation of the radial glial lineage.

Expression of glial fibrillary acidic protein by differentiated astrocytes is regulated by serum antagonistic factors

We report here that, in culture, the expression of glial fibrillary acidic protein (GFAP) by astrocytes, as well as their shape (flat-polygonal vs. stellate) can be regulated by 4 serum antagonistic factors. Three of these factors are stimulatory, while the fourth exerts an inhibitory effect upon these astrocytic properties. As suggested by temperature and trypsin treatments, the inhibitory factor is a polypeptide or a protein of 15-35 kDa. The stimulatory factors are smaller: two of them have a mol. wt. between 0.2 and 5 kDa; the third is smaller than 0.2 kDa. Treatments with chloroform/methanol, ammonium sulfate, neuraminidase, and papain, indicate that at least one glycolipid and one glycoprotein are involved. We speculate that, during development, cells from the astrocytic line could be susceptible selectively to one or another of these factors, which would explain their great plasticity.

Experimental inhibition of reactive gliotic-like changes in astrocytic cultures

Dibutylcyclic AMP (DiBucAMP) can induce astroblast-containing cultures to form cells which resemble reactive astrocytes observed in vivo. In the present study, myelin- or axolemmal-enriched fractions were assessed for the ability to inhibit (DiBucAMP)-stimulated reactive-like changes in astrocytic cultures. Addition of exogenous myelin- or axolemmal-enriched fractions to DiBcAMP-exposed cultures prevented drug-induced elevation of lactic dehydrogenase (LDH) 2 days after initial addition of the drug and moderated DiBucAMP-induced decreases in the enzyme's activity normally observed 5 days later. The proportion of reactive colonies (i.e., those in which more than 50% of cells are stellate-shape) was significantly lower in cultures exposed to the drug plus myelin or axolemma vs those exposed to DiBucAMP alone. The inhibitory factors in axolemmal fractions were heat sensitive (at 100 degrees C), whereas those in myelin were not. Both fractions were inactivated by trypsin. Whole brain homogenates had no effect on diBucAMP-stimulated changes in culture.

The effects of methylmercury on the cytoskeleton of murine embryonal carcinoma cells

Immunofluorescence staining with antibodies to tubulin and vimentin and staining with phalloidin have been used to examine the effects of methylmercury on the cytoskeleton of embryonal carcinoma cells in culture. Exposure of embryonal carcinoma cells to methylmercury (0.01 to 10 microns) resulted in concentration- and time-dependent disassembly of microtubules in interphase and mitotic cells. These effects were reversible when cultures were washed free of methylmercury. Spindle microtubules were more sensitive than those of interphase cells. Spindle damage resulted in an accumulation of cells in prometaphase/metaphase, which correlated with a temporary delay in the resumption of normal proliferation rate upon removal of methylmercury. Of the interphase cytoskeletal components, microtubules were the first affected by methylmercury. Vimentin intermediate filaments appeared relatively insensitive to methylmercury, but showed a reorganization secondary to the microtubule disassembly. Actin microfilaments appeared unchanged in cells showing complete absence of microtubules. Our results 1) support previous reports suggesting that microtubules are a primary target of methylmercury, 2) document a differential sensitivity of mitotic and interphase microtubule systems and 3) demonstrate the relative insensitivities of other cytoskeletal components.

Acquisition of vimentin in astrocytes cultured from postnatal rat brain

Vimentin and glial fibrillary acidic protein (GFAP) represent the principal constituents of intermediate filaments found in astrocytes. In contrast to vimentin-GFAP transition which occurs during glial development in situ, vimentin coexists with GFAP in cortical astrocytes allowed to differentiate in culture. To examine whether culture conditions or proliferative activity of the cells is responsible for the expression of vimentin, we generated cultures of GFAP-positive, vimentin-negative astrocytes isolated from 26-day postnatal rat brain cortices. Isolated astrocytes are characterized by a very thin rim of perinuclear cytoplasm and by numerous processes. Antiserum to GFAP labelled major processes and cell somata of some astrocytes, especially those with relatively short and large processes. Within 3 days in culture, all astrocytes accumulated GFAP in hypertrophic cell bodies and many began to express vimentin. Vimentin appeared primarily close to nuclei, and filaments of vimentin extended into proximal segments of the cell processes. In some astrocytes, however, vimentin was always absent. Combined double immunolabelling and histoautoradiography experiments demonstrated that the acquisition of vimentin was independent of the ability of astrocytes to incorporate tritiated thymidine. The results indicate that astrocytes isolated from 26-day postnatal rat brain are heterogeneous with respect to their ability to express vimentin and that vimentin synthesis is not correlated with the growth state of the cells as had been previously suspected.

The disappearance of a cyclin-like protein and the appearance of statin is correlated with the onset of differentiation during myogenesis in vitro

We have used monoclonal antibodies to statin (S-44) and a cyclin-like protein (S-132) to examine the distribution of these two antigens in proliferating and in nonproliferating populations of cells. We have found that this cyclin-like protein is present in proliferating fibroblasts, whereas statin is absent from these same cell populations; in contrast, in senescent populations of fibroblasts the cyclin-like antigen disappears and statin labeling of nuclei appears. During myogenesis in rat muscle cell cultures, S-132 labeling is present in proliferating myoblasts and disappears after cells fuse and differentiate as multinucleated myotubes. In contrast, statin is absent from proliferating myoblasts, but appears when these cells become postmitotic and begin to differentiate. Similar results were seen during chick myogenesis. We have also found similar results during serum-starvation-induced differentiation in neuroblastoma cells. These results indicate that the cyclin-like protein disappears and statin appears upon commitment to differentiation in vitro, and the presence or the absence of these proteins appears to provide cellular markers for the transition from the proliferative to the nonproliferative state during differentiation.

48-Kilodalton intermediate-filament-associated protein in astrocytes

We provide evidence that a protein of 48 kilodaltons (KD), recognized by a normal rabbit serum (F2N), is associated with intermediate filaments (IF) of astrocytes both in cell cultures and in situ. Immunofluorescence staining shows that the F2N serum gives a fibrous staining pattern similar to that seen with anti-serum to glial filament protein (GFP), a protein specific for IF of astrocytes, and that both proteins are present in the perinuclear fibrous aggregates of IF produced by treating the cells with colchicine. At the ultrastructural level the gold particles decorating the 48-KD protein are localized in clusters along the IF, whereas the gold particles decorating the GFP are localized on the IF in a linear pattern. This difference in distribution and the fact that the two proteins have different electrophoretic mobilities on SDS gels indicates that the 48-KD protein although associated with IF is different from GFP. The 48-KD protein appears to be a distinct, developmentally regulated intermediate-filament-associated protein (IFAP), different from other IFAPs reported to date and the first IFAP described in astrocytes. Its appearance in late developmental stages when motile astroblasts are changing into nonmotile stellate cells suggests that the 48-KD protein may be involved in cross-linking the GFP-containing IF to provide more tensile strength to the cytoplasm at the expense of flexibility.

Ependymoblastoma: a histological, immunohistological and ultrastructural study of five cases

Five ependymoblastomas were studied by means of routine histological techniques, immunohistology and electron microscopy. The tumours were characterized histologically by medium sized, poorly differentiated cells with round or oval nuclei frequently in mitosis and by ependymoblastic rosettes. Reactions for cytokeratin and neurofilament were negative and tubular material positive for glial fibrillary acidic protein was scanty. All five tumors demonstrated positivity for vimentin and S-100 protein. Electron microscopy showed poorly differentiated cells with high nucleo-cytoplasmic ratio and scanty cytoplasmic organelles. Sparse rosettes were present and the cells were united by junctional complexes. Frequent rudimentary or incomplete cilia, a few basal bodies and a few short intercellular glial-like filaments were seen. Features differentiating ependymomas and anaplastic ependymomas from ependymoblastomas are discussed and the need for a definite category separating ependymoblastomas from the former tumours is emphasized.

Oligodendroglial and astroglial heterogeneity in mouse primary central nervous system culture as demonstrated by differences in GABA and D-aspartate transport and immunocytochemistry

Using simultaneous autoradiography and immunofluorescence we have investigated the functional heterogeneity amongst oligodendrocytes and astrocytes in primary mouse central nervous system (CNS) culture as expressed by differences in their ability to accumulate gamma-[3H]aminobutyric acid [( 3H]GABA) and D-[3H]aspartate. We have used a range of specific antibodies that identify oligodendrocytes and astrocytes, from precursor to fully mature cells, to address the question of whether all neuroglial cells are capable of expressing this function. Our results showing that A2B5-, 03-, and galactocerebroside-positive cells became heavily labelled with these two neuroactive amino acids, whereas cells expressing the myelin proteins 2',3'-cyclic nucleotide 3'-phosphohydrolase (CNP) and myelin basic protein (MBP) did not, demonstrate that this capacity is already present in oligodendrocytes at early developmental stages but may not extend to fully mature cells. Astrocytes in culture exhibited a large degree of variability with respect to their ability to transport GABA and D-aspartate. When grown in either serum-containing or serum-free hormone supplemented culture medium two morphologically distinct of glial fibrillary acidic protein (GFAP)-positive astrocyte were identified, process-bearing and epithelioid. Process-bearing cells became heavily labelled with the amino acids under both growth conditions, whereas, data showed that although epithelioid astrocytes were not, or only lightly, labelled with either amino acid in serum-containing cultures, when grown in serum-free culture medium they became more heavily labelled. Thus the expression, in culture, by epithelioid astrocytes, of one of the functions attributed to these cells is largely dependent on growth conditions.

Organization of microfilaments in astrocytes that form in the presence of dibutyryl cyclic AMP in cultures, and which are similar to reactive astrocytes in vivo

This paper deals with changes in the arrangement of microfilaments at various stages during the transformation of astroblasts into reactive astrocytes in the presence of dibutyryl 3',5'-cyclic adenosine monophosphate in vitro. When cultures of two-week-old mouse astroblasts are treated with dibutyryl cyclic AMP, drastic changes occur in cell shape and in the organization of microfilaments, resulting in cells that closely resemble reactive astrocytes in vivo. A thick, prominent ring of microfilaments in such cells which stains strongly with 7-nitrobenz-2 oxa-1,3-diazole-phallacidin, delineates the perikaryon. Electron microscope examination showed that the ring is composed of many smaller bundles of microfilaments running parallel to each other. Prominent bundles of microfilaments radiate from the cell body into the cell processes. Based on the observation of intermediate forms, we propose that the microfilament ring may be important in the development of cell processes in reactive astrocytes.

Glial fibrillary acidic protein immunohistochemistry of spinal cord astrocytes after induction of ischemia or anoxia in culture

The effects of ischemia (removal of oxygen and glucose for 4 h) and anoxia (removal of oxygen alone) on astrocytes were studied in dissociated cultures of E14 spinal cord containing both neurons and astrocytes. In addition, a group of cultures was treated with a low Na+, low Ca2+, and high K+ medium during the 4-h ischemic period (ischemia-protected group), a process that protects neurons from ischemic damage under identical conditions. Astrocytes were examined immunohistochemically using glial fibrillary acidic protein (GFAI) antiserum 24 h after insult. Densitometry and statistical analysis (1-way analysis of variance [ANOVA], a priori; 2-tailed Tukey-t, a posteriori) of the digitized images of the somata and processes of astrocytes in the anti-GFAP reacted cultures showed significant differences between the groups; a significant increase (P less than 0.01) in the GFAP-positive reaction in the somata of ischemic astrocytes and a significant decrease (P less than 0.01) in the GFAP-positive reaction in the processes of ischemic, ischemia-protected, and anoxic astrocytes. There were no significant differences in the GFAP immunoreactivity of somata between control, ischemia-protected, and anoxic astrocytes or of processes from ischemic, ischemia-protected, and anoxic astrocytes. These data show that following ischemia cultured astrocytes increase somatic GFAP immunoreactivity compared to all other groups tested whereas the staining intensity for GFAP was decreased in the processes of all three experimental groups compared to controls. Ischemia protection resulted in the absence of the enhancement of somatic GFAP immunoreactivity. The relationship of the astrocytic response and the type of cellular stress is discussed.

Vimentin dynamics during the mitogenic stimulation of mouse splenic lymphocytes

We have used double immunofluorescence and electron microscopy to examine the distribution of tubulin and vimentin during the stimulation of mouse splenic lymphocytes by the mitogen concanavalin A. In unstimulated cells, vimentin forms a filamentous network partially coincident with the radial pattern of microtubules. In stimulated cells, the numbers of microtubules assembled from the centrosome have increased and vimentin is organized as an aggregate located near the centrosome. When these cells enter mitosis, vimentin is arranged into a filamentous cage enclosing the mitotic apparatus. During cytokinesis, the polar centrosomes are observed at a position adjacent to the midbody and vimentin is detected as an aggregate, similar to that seen prior to mitosis, close to the centrosome in each daughter cell. Using several agents, such as colchicine, colcemid, nocodazole, and taxol, which affect microtubule assembly, we have observed that the vimentin system, although closely related spatially to the microtubule complex in lymphocytes, can still reorganize independently as these cells progress through the cell cycle. Throughout mitogenic stimulation in the continued presence of taxol, microtubules are reorganized into a few thick bundles while the vimentin system undergoes a sequence of rearrangements similar to those observed during normal stimulation. These data suggest that vimentin dynamics may be important in the progression of lymphocytes through the cell cycle in response to mitogen.

Effects of triiodothyronine on the development of GFAP-immunoreactivity and CAT-activity in monolayer cultures of embryonal rat forebrain cells

Dose-dependent actions of triiodothyronine were studied in serum-containing primary cultures of embryonal rat forebrain cells. Triiodothyronine at concentrations exceeding the normal medium concentration by 2 and 5 nM failed to cause significant changes in the development of CAT-activity and GFAP-immunoreactivity. In the presence of additional 10 and 20 nM T3, however, CAT-activity was increased and GFAP-immunoreactivity was decreased in comparison to cultures maintained with 'normal' (10% FCS-containing) media.

Glial filaments in the subcutaneous tumors of mouse glioma clones differently expressing glial fibrillary acidic protein. An immunohistochemical and ultrastructural study

Glial filaments contain vimentin and glial fibrillary acidic protein (GFAP). The question of how glial filaments change morphologically according to the expression of vimentin and/or GFAP has remained unclear. In this study, immunohistochemical and ultrastructural examinations were performed on the subcutaneously transplanted tumors of two clones (F6B3 and G10A10) derived from a mouse glioma. F6B3 tumor expressed GFAP and vimentin in large quantities. G10A10 tumor expressed plenty of vimentin but only a little of GFAP. Ultrastructurally, F6B3 tumor contained abundant cytoprocesses in most of which numerous intermediate filaments (IFs) were arranged in a parallel array. On the other hand, only a small number of the processes were seen in G10A10 tumor, which showed a few IFs arranged either randomly or sparsely in the processes. Both tumors commonly had the IFs accompanied by visible sidearms, but there was a difference in that the smooth and firm IFs were confined to part of F6B3 tumor. Thus, the comparison made between the two models presented differences in the content, arrangement and morphology of IFs, as well as in the manner of GFAP expression, suggesting correlation between these differences.

Heterogeneity of a cultured neoplastic glial line. Establishment and characterisation of six clones

Six clones have been established from a tumorigenic glial cell line (VMDk P497) originally derived from a spontaneous mouse astrocytoma. The clones express dissimilar morphological, antigenic, kinetic and chromosomal properties, thereby indicating the heterogeneity of the parent culture line.

Glial fibrillary acidic protein (GFAP) in oligodendrogliomas: a reflection of transient GFAP expression by immature oligodendroglia

Fourteen pure oligodendrogliomas were studied by light- and electronmicroscopy and immunohistochemistry to examine glial fibrillary acidic protein (GFAP) positivity in the tumors. To compare the immunohistochemical staining patterns of neoplastic oligodendroglia and immature oligodendroglia, myelination glia in the white matter of eight normal brains from children under 6 months of age were studied. The tumors possessed light microscopic and ultrastructural features characteristic of oligodendrogliomas. Microtubules were found in the cytoplasm of nine tumors on electronmicroscopy. In one, intermediate filaments and microtubules were observed in occasional tumor cells with polygonal crystalline structures in the cytoplasm. Using the peroxidase-antiperoxidase technique, all specimens were stained for GFAP, vimentin, S-100 and neuron-specific enolase (NSE). In nine tumors, variable numbers of cells with an oligodendroglial morphology reacted positively for GFAP. All tumors were positive for S-100 and negative for vimentin and NSE. The myelination glia in the eight normal brains stained positively for GFAP but not for vimentin. Vimentin is expressed by developing, reactive and neoplastic astrocytes. Thus, GFAP positivity combined with vimentin negativity in both neoplastic and immature oligodendroglia suggests that GFAP positivity in oligodendrogliomas may reflect the transient expression of this intermediate filament by immature oligodendroglia.

Selective uptake of neuroactive amino acids by both oligodendrocytes and astrocytes in primary dissociated culture: a possible role for oligodendrocytes in neurotransmitter metabolism

CNS glia may be involved in the modulation of neuronal excitability through their capacity to accumulate and metabolize neuroactive amino acids. To investigate the possible role of oligodendrocytes in amino acid neurotransmitter metabolism, we have used light microscopic autoradiography, following the uptake of 3H-labelled amino acids by dissociated cultures of neonatal mouse brain, characterized immunocytochemically using cell-type specific markers. Oligodendrocytes, recognized by their characteristic galactocerebroside membrane staining, rapidly accumulated [3H] gamma-aminobutyric acid (GABA), becoming intensely labelled over cell body and processes after short incubations. In contrast, oligodendrocytes became only lightly labelled with [3H]L-glutamate and aspartate, which preferentially labelled astrocytes. [3H]D-aspartate, a non-metabolized analogue of L-glutamate, was avidly accumulated by oligodendrocytes, labelling cell bodies and processes after short incubations, to a similar extent as GABA. Thus, oligodendrocytes possess a transport mechanism for these excitatory amino acids, but rapidly metabolize them and release the metabolites. Not only the GC-positive cells but also the GC-negative undifferentiated oligodendrocyte precursors accumulated both GABA and D-aspartate, suggesting that this may be a function expressed early in the differentiation of oligodendrocytes. Net uptake of [3H] beta-alanine and [3H]glycine by oligodendrocytes was not observed under any conditions tested. A small number of oligodendrocytes were labelled with [3H]taurine after longer incubations. The uptake of certain neuroactive amino acids is thus a property shared by astrocytes and oligodendrocytes, the latter acting in a protective fashion around neuronal perikarya and axons.

Two simian virus 40 (SV40)-transformed cell lines from the mouse striatum and mesencephalon presenting astrocytic characters. III. A light and electron microscopic study

In two preceding papers we described the cloning of two astrocytic cell lines by simian virus 40 (SV40) transformation of embryonic mouse mesencephalon (F7-Mes) and striatum (F12-Str). The characterization of these lines as belonging to the astrocytic lineage is based on pharmacological, immunocytochemical and physiological data. Here we present quantitative and qualitative data on the morphological aspects of these two astrocytic clones observed under light and electron microscopy. We show that the clones present ultrastructural characters reminiscent of the morphology of young astrocytes. On one hand, they are rather similar to primary astrocytes in culture; on the other, they differ both from a clonal fibroblastic cell line (BT2) and from embryonic mouse fibroblasts in primary culture. These astroblastic clones display 4 morphologically different cell populations which we called types I, II, III and IV. Types II and III are very similar and represent the most predominant cells; their morphologies strongly remind of that of astroblasts. Type I corresponds to glioblasts and does not account for more than 15-20% of the total population. Type IV, which is very similar to differentiated velamentous astrocytes, normally represent ca. 5% of the cells. However, when the transformed cells are treated with mitomycin or mitomycin + dibutyryl cyclic AMP (dbcAMP), the proportion of type IV cells increases very much (up to more than 50% of the cells) while types I, II and III become less numerous. Morphological analysis therefore confirms that the two cell lines derived from the SV40 transformation of 14-day-old embryonic mesencephalic and striatal cells belong to the astrocytic lineage. Moreover, it seems that they can differentiate in vitro in cell culture conditions either spontaneously or under the action of pharmacological treatments known to enhance normal astrocyte maturation.

Immunohistochemical demonstration of vimentin in human cerebral tumors

The distribution of vimentin (VIM) has been histochemically investigated in 53 cerebral tumors and compared in gliomas to that of glial fibrillary acidic protein (GFAP). In gliomas VIM is less positive than GFAP, but shows the same distribution. It cannot be considered as indicating immaturity of glial tumor cells. VIM is also positive in glial processes of cerebellar pilocytic astrocytomas, in Schwann cells of neurinomas and in endothelial cells of all oncotypes. In medulloblastomas, VIM decorates reactive glia cells. A diffuse positive reaction has been observed in meningiomas. In hemangioblastomas, besides intervascular and endothelial cells, groups of polygonal cells are intensely positive for both VIM and GFAP. The interpretation of VIM in cerebral tumors is largely based on the distribution patterns of this intermediate filament in the developing CNS of rodents.

Assembly of glial intermediate filament protein is initiated in the centriolar region

Assembly of glial intermediate filament protein (GFP) into intermediate filaments (IF) was first detected by immunofluorescence in the perinuclear region of astrocytes differentiating in colony cultures before the rest of the cytoplasm was labeled. Double labeling with antisera specific for centrioles indicated that this site corresponds to the centriolar region. These studies suggest that the centriolar region plays an important role in the assembly of some types of IF as well as microtubules.

Primary cultures from defined brain areas; effects of seeding time on cell growth, astroglial content and protein synthesis

The influence of seeding time on cell growth, astroglial content and on protein synthesis during cultivation was determined in primary cultures from 3 phylogenetically different brain areas from rat cerebral cortex, striatum and brainstem. Brainstem cultivated from 17-day-old embryos and all the cultures studied from the 3 brain areas of newborn and 7-day-old rat showed a similar increase in total and water-soluble protein during cultivation. Glial fibrillary acidic protein (GFAp, alpha-albumin) levels increased with age in all cultures studied. There was a rapid increase in GFAp (alpha-albumin) between 1 and 2 weeks in cultures from newborn and between 2 and 3 weeks in brainstem cultures from 17-day-old embryos, these increases being slower thereafter. Incorporation of [3H]valine into soluble protein was lower in 3-week-old cultures than in 1- and 2-week-old cultures derived from newborn and 7-day-old rat brain. The incorporation rates were similar in comparisons of the various cultures. Similar results were obtained from embryonic cultures, although the decrease in incorporation rate was between 3 and 4 weeks. The efficiency of incorporation (% TCA-precipitated material/total [3H]activity) was higher in 2- and 3-week-old than in 1-week-old cultures from newborn and 7-day-old rats and in 3- and 4-week-old cultures of brainstem from 17-day-old rat embryos. These findings suggest a cell differentiation during cultivation. The results show that seeding time has a variable influence on cultures from the different brain areas studied concerning cell growth, astroglial content and probably differentiation during cultivation. Embryonic cell cultures seem, in general, to develop one week later than neonatal and postnatal ones. Cultures of newborn rat cells from cerebral cortex, striatum and brainstem show many similarities in the above parameters during cultivation. This is also the case for brainstem cultures from embryonic rat.

Astrocyte cell lineage. IV. Changes in the organization of microfilaments and adhesion patterns during astrocyte differentiation in culture

The organization of microfilaments using NBD-phallacidin and cell adhesion to substratum by surface reflection interference microscopy was examined during differentiation of astrocytes in colony cultures and correlated with motile behaviour of cells. Disaggregated cells from neopallium of 12-day-old or newborn DBA/1J mouse embryos were used to establish colonies and astrocyte precursor cells at various stages of differentiation along the astrocyte lineage were examined after 3 days, 1, 2 and 4 weeks in culture. The earliest astrocyte precursor cells, the glioblasts, are stationary and form epithelial-type colonies which adhere to the substratum primarily around the edge where large microfilament bundles are found. Bundles of microfilaments are also present around the apical ends of closely packed cells. As the epithelial cells start to separate and transform into flat proastroblasts, adherens-type junctions which have a zig-zag appearance and are associated with microfilaments form between adjacent cells. In the highly motile astroblasts these junctional regions break down into multiple smaller regions where the separated cells remain in contact through fine processes. The astroblasts also have stress fibres, focal contacts with substratum, foci from which microfilament bundles radiate and a complex pattern of fine, circumferentially oriented bundles of microfilaments. This elaborate organization of microfilaments disappears as the motile astroblasts differentiate into stationary fibrous astrocytes that have little polymerized actin and lack focal contacts. These results show that stationary astrocyte precursor cells in vitro go through a highly motile stage having a characteristic distribution of microfilaments and focal contacts before becoming stationary again. We consider that the motile stage could correspond to the stage in vivo when astrocyte precursor cells migrate from the ventricular and subventricular regions to take up position in different parts of the developing brain.

Interleukin-1 stimulation of astroglial proliferation after brain injury

The interleukins, which have a regulatory role in immune function, may also mediate inflammation associated with injury to the brain. In experiments to determine the effect of these peptide hormones on glial cell proliferation in culture, interleukin-1 was a potent mitogen for astroglia but had no effect on oligodendroglia. Interleukin-2 did not alter the growth of either type of glial cell. Activity similar to that of interleukin-1 was detected in brains of adult rats 10 days after the brains had been injured. These findings suggest that interleukin-1, released by inflammatory cells, may promote the formation of scars by astroglia in the damaged mammalian brain.

Distribution of glutamine synthetase and glial fibrillary acidic protein and correlation of glutamine synthetase with glutamate decarboxylase in different regions of the rat central nervous system

The concentration of soluble glial fibrillary acidic (GFA) protein and the specific activity of glutamine synthetase (GS) were estimated in 11 central nervous system (CNS) regions of the 90-day-old rat. Marked differences were observed in the regional distribution of these astrocyte marker proteins. The striatum and spinal cord contained the lowest concentration (per g wet weight) of GFA protein and GS activity, respectively, while the olfactory bulbs had the highest level of both astrocytic proteins. Differences between the lowest and the highest values were 3-fold for GS and 4-fold for GFA protein. More significant was the marked variation in the ratio of GS to GFA protein in different CNS regions; the highest and lowest values were in the striatum and the spinal cord respectively, and the difference between the highest and the lowest value was about 5-fold. The spinal cord contained low GS and high GFA protein; on the other hand, the colliculi had high GS and relatively low GFA protein. Immunochemical detection of GS and GFA proteins in whole homogenates of different regions showed that the variation of the specific activities of GS and the concentration of soluble GFA protein were due to the differences in their absolute protein concentrations. In different regions of the brain the activity of GS was significantly correlated with that of glutamate decarboxylase, but not with that of choline acetyltransferase. These observations provide further evidence for differing biochemical properties of astrocytes from various CNS regions and for the involvement of GS in processes associated with amino acid neurotransmission.

Effect of cytosine arabinofuranoside (AraC) on reactive gliosis in vivo. An immunohistochemical and morphometric study

Glial reactivity is believed to contribute to the lack of functional recovery after injury to the mammalian central nervous system. The role of glial mitosis in the progression of events associated with reactive gliosis has received little attention. In the present study, the expression of reactive gliosis distal to the site of crush in rat optic nerves was assessed in the presence and absence of a chronically administered mitotic inhibitor, cytosine arabinofuranoside (AraC). Right optic nerves were crushed and animals sacrificed 1, 2 or 3 weeks later. Parameters assessed were (1) glial hypertrophy, (2) degradation of myelin sheaths and (3) ability of tissue to stain with antisera raised against glial filament protein (GFA), actin and vimentin. In saline treated animals, greater than 90% of the myelin sheaths distal to the site of axonal injury had degraded within 7 days postoperatively. Glial hypertrophy was evident by the second week after after crush and increased progressively. The number of GFA-positive profiles (i.e., cells) increased between 1 and 3 weeks postoperatively. Vimentin staining increased 4-fold between 1 and 2 weeks after injury and subsequently showed no change. Actin staining rose 3-fold between 1 and 2 weeks after injury, but decreased by the third postoperative week. In AraC treated animals, almost 50% of the myelin sheaths distal to the injury site were preserved a week after surgery. A delay in myelin degradation continued until the second postoperative week. Glial hypertrophy was evident at the 2 and 3 week time points. However, the extent of hypertrophy was substantially lower in drug (vs saline) treated animals. Vimentin staining never rose above minimal levels in AraC treated animals. Actin staining in AraC rats at 2 weeks postoperatively was equivalent to that in saline injected animals, but in contrast to the results in the latter group, increased (3-fold) between 2 and 3 weeks after crush. Results indicate a delay in the expression of reactive gliosis with chronic administration of AraC. It is proposed that this might be due to a delay in the appearance of 'signals' (e.g., myelin debris) which initiate the process of reactive gliosis.

Observations on cell growth and regulation of glutamine synthetase by dexamethasone in primary cultures of forebrain and cerebellar astrocytes

Cell growth, development of glutamine synthetase and its regulation by glucocorticoids, were studied in primary cultures of two types of astrocytes derived from rat brain, one from newborn forebrain and another from either newborn or 8-day-old cerebellum. Cell number per dish increased linearly following an initial decrease due to removal of non-astrocytic cells, and after the second week reached a stationary phase in both types of cultures at more or less the same time, although the cell number in cerebellar cultures was about 35% lower than in forebrain cultures of the same age. At all ages, irrespective of cell density, the cerebellar astrocytes were larger in size than the forebrain astroglial cells. The developmental curves for glutamine synthetase activity were similar in vitro and in vivo; however, the increase in enzyme activity in vitro was significantly greater than in vivo and this difference was more marked in forebrain than in cerebellar cultures. Throughout the period studied the specific activity of glutamine synthetase was significantly higher in forebrain astrocytes than in cerebellar astroglial cells. Treatment with dexamethasone caused a marked increase in the specific activity of glutamine synthetase. However, in agreement with our previous in vivo findings, the steroid induction in forebrain astrocytes was significantly less than that in cerebellar astrocytes. In culture, both types of astrocytes remain responsive to the hormone for longer than in vivo. The differences in the biochemical properties of the forebrain and cerebellar astrocytes seem to be intrinsic, and not related to the cell density or to the purity of the cultures.

Motoneurone survival is induced by immature astrocytes from developing avian spinal cord

Dissociated spinal cords of 6-day chick embryos were grown on monolayers consisting primarily of either flat (relatively immature) or process-bearing (relatively mature) astrocytes. Cultures rich in flat astrocytes maintained about 80% of the motoneurones originally plated for 48 h in vitro. However, process-bearing astrocytes were unable to support motoneurone survival. Medium conditioned by contact with the monolayers of flat astrocytes also promoted motoneurone survival for 48 h. Maximal activity occurred over the concentration range 55-110 micrograms/ml protein. After 48 h, the number of motoneurones dropped to control levels both in the conditioned medium and on the monolayers. This effect could not be reversed by the introduction of fresh conditioned media at 48 h. This indicated a decrease in the requirements of more mature motoneurones for this media as muscle-conditioned medium could support 80% of the motoneurones initially plated for 96 h. Thus, relatively immature astrocytes were capable of supporting the survival of 6-day montoneurones in vitro for up to 48 h and this effect is mediated through the release of a soluble substance.

Release of preloaded taurine and hypotaurine from astrocytes in primary culture: stimulation by calcium-free media

The spontaneous and stimulated release of taurine and hypotaurine from astrocytes in primary cultures were investigated. Spontaneous efflux was slow, less than one half of preloaded labeled taurine and hypotaurine still remaining in the cells after a 60-min efflux period. The release processes of both amino acids were in principle similar. No homo- or heteroexchange with extracellularly added taurine, hypotaurine or GABA could be detected, and depolarizing potassium concentrations failed to stimulate taurine or hypotaurine release. On the other hand, omission of calcium ions from medium increased efflux of taurine and hypotaurine about three- and twofold, respectively, in both high-K+ and normal-K+ media.

Surface reflection interference microscopy: a new method for visualizing cytoskeletal components by light microscopy

Surface reflection interference microscopy of detergent resistant residues of cultured cells stained with protein dyes can be used to obtain high resolution images of the cytoskeleton. We have compared the images obtained using different dyes and have examined the effect of four of these dyes on the visualization of different parts of the cytoskeleton in detail. The dependence of contrast of the images obtained on the illuminating numerical aperture and the wavelength of incident light was determined. Staining with Acid Yellow 36, Guinea Green B and Naphtol Blue Black produces images from the entire cytoskeleton and contrast in these images is relatively insensitive to changes in the incident wavelength. Coomassie Brilliant Blue R250 images, on the other hand, result primarily from reflection from the lower surface of the cytoskeleton and the contrast of these images is sensitive to changes in incident wavelength dropping abruptly in the region of the transmission peak of the stain. From the different spectral sensitivities of the reflection images obtained and from differential interference effects at low and high illuminating numerical apertures, we conclude that the reflection images obtained using the first three stains result from modulation of the reflection by interference effects. In contrast, in the case of Coomassie Brilliant Blue R250 the resulting image originates mainly from selective reflection of wavelength near the absorption range of the dye.

Isolation and transplantation of oligodendrocyte precursor cells

Newborn DBA/1J mouse neopallium was disaggregated and grown in high cell densities in tissue culture. In culture, the oligodendrocyte cell precursors are recognized as small refractile cells which use astrocyte precursor cells as a substratum. Using metrizamide density gradients, the oligodendrocyte precursor cells were separated from the astroblasts after 7 days in culture and then transplanted into the cerebellums of neonatal mice. The differentiation of the cultured oligodendrocyte precursors was analyzed in the transplants by nuclear morphometry, light and electron microscopy and immunocytochemistry. Analysis of the experiments indicated that the oligodendrocyte precursor cells, initially grown in culture, differentiated and myelinated host neuronal processes after transplantation. Moreover, the ultrastructure of the transplanted oligodendrocytes resembled mature oligodendrocytes in situ.

Astrocyte cell lineage. III. The morphology of differentiating mouse astrocytes in colony culture

Disaggregated cells of newborn DBA/1J mouse neopallium were grown in colony cultures, and colonies of cells at various stages of differentiation along the astrocyte cell lineage were examined after 3 days, 1, 2 and 4 weeks by electron microscopy and by NBD-phallacidin which demonstrates the distribution of microfilaments. The earliest astrocyte precursor cells or glioblasts are closely apposed epithelial cells that rarely have junctions. Their scanty cytoplasm contains many free ribosomes but few microfilaments. The cells in the next stages of astrocyte lineage or proastroblasts are flat and are separated from each other to a variable degree. They have intercellular junctions associated with microfilaments and contain singly dispersed intermediate filaments. The proastroblasts gradually differentiate into astroblasts which have a similar morphology except that in addition to the singly distributed intermediate filaments they also contain intermediate filaments arranged into bundles of various sizes. The mature fibrous astrocytes have well-defined processes and distinct perikarya. They form from astroblasts in culture and also contain numerous bundles of intermediate filaments. The dibutyryl-cyclic AMP (dBcAMP)-induced astrocytes in culture in contrast are large stellate cells similar to reactive astrocytes found around sites of injury in the brain. On the basis of these and previous immunocytochemical studies of the formation and distribution of intermediate filaments in the cytoplasm of differentiating astrocytes, criteria are proposed for identification of different cells along the astrocyte lineage.

Proteolytic degradation of vimentin and glial fibrillary acidic protein in rat astrocytes in primary culture

Vimentin and glial fibrillary acidic (GFA) protein in rat astrocytes in primary culture were identified with their respective specific antisera on electroblots of one-dimensional and two-dimensional polyacrylamide gels. Each antiserum revealed respectively vimentin (58 kDa, pI 5.3), GFA protein (50 kDa, pI 6-5.8) and several more acidic, lower molecular mass vimentin-immunoreactive and GFA protein-immunoreactive polypeptides. These were more apparent when astrocyte proteins were extracted in the absence of EGTA and in the presence of Ca2+ ions suggesting that the lower immunoreactive polypeptides result from a Ca2+-sensitive proteolytic degradation of vimentin and GFA protein. The extent of proteolysis of GFA protein was higher when the differentiation of astrocytes was induced with 1,N6-dibutyryl-adenosine 3',5'-monophosphate (Bt2cAMP). In contrast, the proteolysis of vimentin was not affected by the state of differentiation of astrocytes. The results indicate the existence of at least two distinct degradative pathways for intermediate filament proteins in astrocytes in primary culture and suggest that in differentiated astrocytes the activity of proteinase(s) involved in the degradation of GFA protein increases.

Astrocyte cell lineage. V. Similarity of astrocytes that form in the presence of dBcAMP in cultures to reactive astrocytes in vivo

The relationship between astrocytes forming in the presence of dibutyryl cyclic AMP (dBcAMP) in culture and reactive astrocytes responding to a cerebral cortex stab wound was investigated using computerized image analysis (Zeiss IBAS 1) and immunocytochemical staining. The diameters of the nuclei of astrocytes in primary cultures of newborn mouse neopallial cells were compared to those of the nuclei of normal and reactive astrocytes in histological sections of mouse cerebral cortex. We found that the nuclei of astrocytes that formed in the presence of dBcAMP in cultures are significantly larger than those of spontaneously occurring small stellate astrocytes in culture and of normal astrocytes of the cerebral cortex in vivo but corresponded more closely to the nuclei of reactive astrocytes in the area surrounding a stab wound in the cerebral cortex. Large stellate cells formed in the presence of dBcAMP had vimentin and an increase in GFP-containing intermediate filaments. Formation of reactive astrocytes in vivo is also associated with an increase in both vimentin and GFP-containing intermediate filaments. These observations indicate a closer relationship of astrocytes formed in the presence of dBcAMP in cultures to the reactive astrocytes in the cerebral cortex than to normal astrocytes. We propose, therefore, that the large stellate astrocytes that form in the presence of dBcAMP be referred to as reactive astrocytes in culture.

Immunocytochemical characterization of cloned cells from normal adult rat brain

A cell clone, isolated from normal adult rat brain and maintained in culture for many passages, has been previously characterized by electron microscopy. Immunohistochemistry has now been shown that it possesses glutamine synthetase (GS) which supports its astrocytic origin. Cells failed, however, to express glial fibrillary acidic protein (GFAP) but did express the intermediate filament protein vimentin. The presence of fibronectin confirms the normal derivation of these cells and excludes the possibility of malignant transformation.

Temporal relationship between the appearance of vimentin and neural tube development

Intermediate filaments of the vimentin type that were initially identified within mesodermally derived cells have recently been demonstrated within several immature cell types derived from neuroectoderm, such as astroblasts and early stage neuroblasts. The objective of the present study was to determine the earliest developmental stage at which vimentin could be detected in the mouse neural tube. Vimentin was not detectable in the newly formed neural tube in E8 embryos. In the E9 neural tube the first positively labeled processes were observed in the ventrolateral region of the cervical neural tube with the processes having the distribution and appearance of those of radial glial cells. Between E9 and E10 there was a significant increase in the vimentin content of the neural tube as labeled filamentous bundles were observed throughout the ventricular cell layer and in the forming mantle layer. The distribution of labeled filaments in the E11 neural tube was similar to that of the E10 tissue although staining intensity was greater in the mantle layer in the E11 tissue. This work identifies the temporal relationship between the appearance of vimentin and neural tube development.