Journal of Biotechnology, 16 (1990) 221-232 Elsevier
221
BIOTEC 00548
Selective adherence of neurons and glial cells from dissociated cerebral and spinal cord microcarrier cultures A. Shahar
1, S. R e u v e n y 1 y .
David 1, G. H a m d o r f 2, M. T e r b o r g 2 and J. Cervos-Navarro 2
I Section of Electron Microscopy, Department of Virology, Israel Institute for Biological Research, P.O. Box 19, Ness-Ziona 70450, Israel," 2 Institute of Neuropathology, Free University of Berlin, Berlin, F.R.G. (Received 25 July 1989; accepted 28 March 1990)
Summary In stationary cultures of dissociated brain and spinal cord grown on microcarriers (MCs), the neuronal and ependymal cells attached to the MCs forming floating aggregates in which they grow in a three-dimensional pattern. The glial and meningeal elements on the contrary, tend to dissociate from the aggregates and adhere to the plastic dish where they divide to form a monolayer. This different behaviour of CNS components is not observed in rotating cultures in which all CNS cells remain attached to the MCs and develop into mature floating structures. This cell separation in stationary MC-cultures which is documented here by SEM and immunocytochemistry, may be useful for analysis and evaluation of the metabolic biochemical events of each of the cellular components derived from the same culture. Microcarrier; Neurons; Glial cell; Cell separation; CNS; Immunocytochemistry
Introduction
The classical in vitro method for studying function and differentiation of dissociated cells from the central nervous system (CNS), involves growing bidimensional Correspondence to: Dr. A. Shahar, Section of Electron Microscopy, Dept. of Virology, Israel Institute for Biological Research, P.O. Box 19, Ness-Ziona 70450, Israel. 0168-1656/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
222 monolayer cultures on polylysine coated supports (Yavin and Yavin, 1977; Sensenbrenner, 1977). One way of obtaining a three-dimensional growth pattern of CNS cells is by cultivating them on microcarriers (MCs). In these MC-cultures, cells are propagated on small solid particles suspended in the culture medium. Several types of MCs having different forms and sizes are commercially available (Reuveny, 1983). We have cultured dissociated cells of the CNS on positively charged cylindrical microcarriers on which they grow and differentiate in a tridimensional pattern (Shahar et al., 1983; 1984a). Such MC-cultures are grown either in rotating flasks or in stationary plastic dishes (Shahar and Reuveny, 1987; Shahar, 1990). In both these methods, cells attach to the MCs and form aggregates which usually remain floating in the medium during cultivation. In the rotating system, due to the rotation, all types of cultured cells become firmly attached to the MCs which form aggregates. In stationary cultures, during the first period of cultivation, the cells are not yet tightly interconnected, thus when aggregates sediment near the bottom of the dish, a number of glia and flat cells detach from the aggregates and adhere to the plastic. In this way two separate cultures are obtained: in one culture most of the neuronal components remain attached to the MCs of the floating aggregates and grow in a tridimensional pattern. In the other culture, the non neuronal elements containing meningeal and glial cells, as well as fibroblasts, grow in the form of a monolayer on the plastic of the dish. For the identification of the cell types, glial fibrillary acidic protein (GFAP) was employed. It is widely used as an immunohistochemical marker for astrocytic cells (Cordell et al., 1984). Since G F A P was found to be a major constituent of glial fibers, the expression of G F A P has been interpreted as a marker for astroglial histogenesis a n d / o r differentiation. Neuron specific enolase (NSE) has been introduced as a marker of neuronal histogenesis and differentiation by several authors (Pickel et al., 1976; Schmechel et al., 1978). Kumpulainen and Korhonen (1982) reported that mouse and human oligodendrocytes are Carboanhydrase C (CA-C)-positive, and therefore CA-C-antiserum can be used as a marker for normal oligodendrocytes. In vivo results have shown that in rat brain, the most active synthesis of this enzyme occurred during the myelination period. In glial cell cultures, the level of CA increased in parallel to the development of the oligodendroglia-like cells and the enzyme was only localized in these cells (Delaunoy et al., 1989). In this study, we used scanning electron microscopy (SEM) to determine the tridimensional growth of neurons on MCs, and immunocytochemistry for the characterization of glia cells grown on the plastic.
Material and Methods Microcarriers
We have used cylindrical MCs (DE-53) produced by Whatman, U.K. and marketed as an anion exchange resin for chromatography. These MCs measure
223 80-400/~m in length and 40-50/~m in diameter. They are made of cellulose matrix, charged with tertiary amine groups. For use, they are equilibrated with phosphatebuffered saline (pH 7.4) and autoclaved (121 ° C, 20 min) in batches of 5 g in 100 ml. After being washed twice in nutrient medium. MCs are added to cultures to a final concentration of 3-5 mg per 1 0 6 ceils. Since MCs tend to stick to glass, they should be handled with either plastic or siliconized glass instruments. For further details on the use of MCs, see Reuveny (1983) and Shahar (1990).
MC-cultures Brain and spinal cords were dissected, from Sprague-Dawley rat fetuses at the ages of 16 and 14 d, respectively. Cells were obtained by mechanical dissociation (Yavin and Yavin, 1977) and were suspended with the MCs as follows: For rotating cultures, 8-10 x 1 0 6 cells and 25-30 mg of MCs were suspended in 50 ml Erlenmeyer shake flasks containing 10 ml of nutrient medium. Cultures were rotated at 150 rotations per min (1 inch amplitude). For stationary cultures 1 x 1 0 6 cells and 3 mg of MC cells were plated in 32 mm plastic dishes without substrate coating, in 1.5 ml of nutrient medium per dish. The culture medium for brain cells consisted of 20% fetal calf serum and 77% Eagle's basal medium. The spinal cord culture medium contained 10% of each fetal calf serum and horse serum and 77% of MEM-Eagle's. Both media were supplemented with 1% of 2 mM L-glutamine, 1% glucose (from a 50% solution to provide a final concentration of 600 mg) and of 1% 16/~g ml -a gentamicin solution. The sera were heat inactivated at 56 o C for 25 min. Stationary cultures were incubated in a CO2 incubator and their medium was changed twice a week. On the second, and sometimes also on the third feeding of the stationary cultures, the aggregates were carefully transferred to new uncoated plastic dishes, using a curved Belco pipette. The remaining dishes, with the attached cells, were cultured for a few more days prior to fixation. The rotating cultures were pregassed with 5% CO 2 in air and closed with rubber stoppers. Only 50% of the medium was replaced twice a week.
Processing of aggregates for SEM Aggregates were collected in conic tubes and fixed in 2.5% glutaraldehyde, postfixed in 1% osmium tetroxide and dehydrated in increments of ethanol. Critical point drying was made in a Polaron apparatus using liquid CO 2. Aggregates were then placed on double scotch tape and vacuum coated with 200 A gold-palladium (Polaron sputtering unit). Observation was made in a JSM 35C at 25 KV.
Immunocytochemistry G F A P for astroglial cells was used in the alkaline phosphatase-anti alkaline phosphatase (APAAP) technique according to Cordell et al. (1984). Dilution of GFAP-antibodies was 1 : 250 (Dakopatts, Denmark). NSE as a neuron specific marker and CA-C as a marker of oligos were used according to the techniques described by Nakagawa et al. (1986). The dilution of anti-neuron-specific enolase was 1:1000 (Dakopatts, Denmark) and that of anticarbonic anhydrase C was 1 : 50 (Calbiochem-Behring).
Fig. 1. SEM of MC-aggregates 15 d in vitro (DIV). A. Flat cells exhibiting numerous microvilli in spinal cord cultures. B. Perikarya and a network of interconnecting fibers of brain cells.
225 Results
In rotating CNS cultures, mainly of spinal cord, in which meninges had not been stripped off, most of the aggregates during the entire period of cultivation were composed of a mixed cell population. The cells were composed, in certain areas, of round or fusiform perikarya, which formed a densely packed network of nerve fibers among them (Fig. 1B). Other areas were covered with a sheath of adjacent fibroblasts and meningeal ceils exhibiting numerous microvilli. With time, the non-neuronal elements divided and ensnared most of the surface of the aggregates (Fig. 1A). On the other hand, in stationary cultures, due to detachment of a large number of glial cells, meningeal elements and fibroblasts, the aggregates were composed mainly of nerve cells, which adhered directly to the MCs and sprouted to form a well organized fiber network (Figs. 2, 3). In some aggregates, mainly of brain cultures, several parts were covered by ependymal cells which remained attached to the MCs and formed epithelial like laminae (Fig. 4A). The luminal surface of the ependymal cells was convex and covered with numerous microvilli, many of the cells exhibited groups of cilia (Fig. 4B). Aggregates which were formed in spinal cord MC-cultures contained, in addition to neuronal cells, also elongated myotubes which were arranged adjacent and parallel to each other (Shahar et al., 1986). A few aggregates in stationary cultures might attach to the plastic of the dish and act as explants with nerve fibers sprouting over a layer of migrating non neuronal elements (Fig. 5A). In spinal cord MC-cultures a few contracting multinucleated myotubes can be observed in the migration zone. However, beside these single attached aggregates with their migration cell zones, the entire plastic dish, in stationary MC-cultures, was mainly covered by a confluent monolayer made of detached cells. The predominating glial cell type in the confluent monolayer culture was considered to be astroglial because the bodies of these cells, and their processes were intensively stained with labelled antibodies against GFAP. These cells were of the average size of 1 5 - 2 0 / , m and were either flat or spindle-shaped with a few processes. Their unstained central oval nuclei contained usually one or two nucleoli (Fig. 5B). The other glial cell type which can be recognized in the monolayer cultures is the oligodendrocyte. These cells appear in considerable numbers and stain positively with anti-carboanhydrase (Fig. 6A). They are small cells with a central prominent nucleus surrounded by a thin cytoplasmic area. Cytoplasmic processes are not always recognized. In the monolayer, there were no NSE positively stained neurons (Fig. 6B). However, in the vicinity of attached aggregates, sometimes single neurons could be observed. Cells in the background which do not stain with immunocytochemical dyes are flat meningeal cells, fibroblasts, as well as round refractile macrophages. Only during the first week of cultivation (prior to the first and perhaps the second transfer of the neuronal aggregates), glial elements and flat cells detached to form a monolayer. During the following weeks in culture, usually no more non-neuronal elements detached.
226
227
Fig. 3. SEM showing two neurons and their processes attached to MCs in the aggregate. Note some flat cells remaining in A. Brain cultures, 20 DIV.
Fig. 2. Intensive network of neuronal fibers and a few rounded and flattened perikarya which remain in the aggregates. Note in A and B the peculiar growth of nerve fibers directly attached to the MCs--unlike in monolayer cultures in which they usually sprout above a layer of dividing non neuronal elements. In C elongated nerve fibers ramify to interconnect with cells which are attached to distant MCs within the same aggregates. SEM, brain cells, 15 DIV.
228
229
Fig. 5. Light microscopy showing in: A. Cell migrating from an aggregate attached to the plastic. B. Astroglial cells stained with labelled antibodies against GFAP, 9 DIV. Original magnifications: A, x 150; B, x 325.
Fig. 4. SEM of ependymal cells from a 14 d brain culture. Cells are organized on the MCs in a large epithelial layer which is probably formed following division of the ependymal cells. The ceils exhibit groups of cilia and microvilli.
230
Fig. 6. Light microscopy of glial cells detached from stationary aggregates after 9 DIV. A, Oligos stained with anticarbonicanhydrase; B, NSE stain showing absence of neurons in the monolayer. Original magnification x 150.
231
Discussion
We have previously reported on the use of MC-cultures, either in stationary or rotating system, as a tool for the study of CNS cell differentiation in vitro. Dissociated brain cells, when grown on MCs, develop synaptic units and become myelinated (Shahar et al., 1983). In spinal cord MC-cultures, in addition to nerve cell differentiation, there is also growth of muscle fibers which mature to form nerve-muscle interconnections (Shahar et al., 1984, 1987). The aggregates in the stationary nerve-muscle cultures were not transferred during the entire cultivation period and were composed of both neuronal and non-neuronal elements. A similar growth pattern of heterogenous cell population of floating aggregates has been described by others in aggregating CNS cultures maintained in stationary and rotating systems (Seeds et al., 1977; Bjerkvig et al., 1986; Bjerkvig, 1986). Here we describe a different approach for cultivation of dissociated CNS on MCs so as to separate different cellular components into two culture types: 1. The neuronal elements, ependyma (and in spinal cord cultures also muscle fibers) remain attached to the MCs forming aggregates in which they establish in a tridimensional pattern. 2. Following the transfer of the neuronal aggregates, the remaining detached glial, meningeal and other non-neuronal elements develop into a monolayer on the plastic dish. Immunocytochemical stains indicate that the predominant glial cell type in the monolayer are astrocytes, though there is also a considerable number of oligos present. This separation of cellular components by the stationary MC-culture technique has several advantages: In the first place it allows to study the specific effects induced by chemical, physical and biological agents acting on either the nerve and the nerve-muscle cell population or on the glia and meningeal elements. Furthermore, on the metabolic-biochemical level, the secretion of factors, neurotransmitters, metabolites, etc., by either of the cellular components can be analysed and evaluated. Finally, the neuronal or nerve-muscle MC entities are easily transferred to other monolayer cultures of different tissues, or can be implanted into living, injured or dystrophic, tissues.
References Bjerkvig, R. (1986) Reaggregation of fetal rat brain cells in stationary culture system II: Ultrastructural characterization. In Vitro Cell Dev. Biol. 22, 193-200. Bjerkvig, R., Steinsvag, S.L. and Laerum, O.D. (1986) Reaggregation of fetal rat brain cells in a stationary culture system I: methodology and cell identification. In Vitro Cell Dev. Biol. 22, 180-192. Cordell, J.L., Falini, B., Erber, W.N., Ghosh, A.K., Abdulaziz, Z., MacDonald, S., Pulford, K.A.F., Stein, H. and Mason, D.Y. (1984) Immunoenzymatic labelling of monoclonal antibodies using immune complexes of alkaline phosphatase and monoclonal antialkaline phosphatase (APAAP complexes). J. Histochem. Cytochem. 32, 219-229. Delaunoy, J.P., Hog, F., Devilliers, G., Bansart, M., Mandel, P. and Sensenbrenner, M. (1980) Developmental changes and localization of carbonic anhydrase in cerebral hemispheres of the rat and rat glial cell cultures. Cell. Mol. Biol. 26, 235-240. Kumpulainen, T. and Korhonen, L.K. (1982) Immunohistochemical localization of carbonic anhydrase isoenzyme C in the central and peripheral nervous system of the mouse. J. Histochem. Cytochem. 30, 283-292.
232 Nakagawa, Y., Perentes, E. and Rubinstein, J.L. (1986) Immunohistochemical characterization of ohgodendrogliomas: an analysis of multiple markers. Acta Neuropathol. 72, 15-22. Pickel, V.M., Reis, D.J., Marangos, P.J. and Zomzely-Neurath, C. (1976) Immunocytochemical localization of nervous system specific protein (NSP-R) in rat brain. Brain Res. 105, 184-187. Reuveny, S. (1983) Microcarriers for culturing mammalian cells and their application. In: A. Mizrahi and A.L. Van Wezel (Eds.), Advances in Biotechnological Processes, Alan R. Liss Inc., New York, Vol. 2, pp. 1-32. Schmechel, D., Marangos, P.J., Zis, A.P., Brightman, M. and Goodwin, F.K. (1978) Brain enolases as specific markers of neuronal and glial cells. Science 199, 313-315. Seeds, N.W., Ramirez, G. and Marko, M.J. (1977) Aggregates cultures: a model for studies of brain development. In: R.T. Acton and J.D. Lynn (Eds.), Cell Culture and its application, Academic Press, New York, pp. 23-37. Sensenbrenner, M. (1977) Dissociated brain cells in primary cultures. In: S. Federoff and L. Hertz (Eds.), Tissue and Organ Cultures in Neurobiology, Academic Press, New York, pp. 191-213. Shahar, A., Reuveny, S., Amir, A., Kotler, M. and Mizrahi, A. (1983) Synaptogenesis and myelination in dissociated cerebral microcarrier cell culture. J. Neurosci. Res. 9, 339-348. Shahar, A., Amir, A., Reuveny, S., Silberstein, L. and Mizrahi, A. (1984a) Neuronal cultures in microcarriers: Dissociated spinal cord cells. Develop. Biol. Standard 55, 25-30 S. Karger, Basel. Shahar, A., Reuveny, S., Mizrahi, A. and Shainberg, A. (1984b) Differentiation of dissociated embryonic central nervous system cells and of myoblasts cultured on microcarriers. J. Acad. Med. Torino CXL VII, 33-38. Shahar, A. and Reuveny, S. (1987) Nerve and muscle cells on microcarriers in culture. In: A. Fiechter (Ed.), Advances in Biochemical Engineering, Vol. 34. Shahar, A. (1990) Cultivation of nerve and muscle cells on microcarriers. Methods Neurosci. 2, 195-209. Shahar, A., Bidder, M., David, Y., Amir, A. and Shainberg, A. (1987) Autologous neuronal and muscle co-cultures: a model for neurotoxicology. In: A. Shahar and A.M. Goldberg (Eds.), Model Systems in Neurotoxicology. Alternative Approach to Animal Testing, Alan R. Liss Inc., New York, pp. 45-58. Yavin, Z. and Yavin, E. (1977) Synaptogenesis and myelinogenesis in dissociated cerebral cells from rat embryo on polylysine coated surfaces. Exp. Brain Res. 29, 137-147. Yavin, E. and Yavin, Z. (1974) Attachment and culture of dissociated cells from rat embryo cerebral hemispheres on polylysine-coated surface. J. Cell Biol. 62, 540-546.
221
BIOTEC 00548
Selective adherence of neurons and glial cells from dissociated cerebral and spinal cord microcarrier cultures A. Shahar
1, S. R e u v e n y 1 y .
David 1, G. H a m d o r f 2, M. T e r b o r g 2 and J. Cervos-Navarro 2
I Section of Electron Microscopy, Department of Virology, Israel Institute for Biological Research, P.O. Box 19, Ness-Ziona 70450, Israel," 2 Institute of Neuropathology, Free University of Berlin, Berlin, F.R.G. (Received 25 July 1989; accepted 28 March 1990)
Summary In stationary cultures of dissociated brain and spinal cord grown on microcarriers (MCs), the neuronal and ependymal cells attached to the MCs forming floating aggregates in which they grow in a three-dimensional pattern. The glial and meningeal elements on the contrary, tend to dissociate from the aggregates and adhere to the plastic dish where they divide to form a monolayer. This different behaviour of CNS components is not observed in rotating cultures in which all CNS cells remain attached to the MCs and develop into mature floating structures. This cell separation in stationary MC-cultures which is documented here by SEM and immunocytochemistry, may be useful for analysis and evaluation of the metabolic biochemical events of each of the cellular components derived from the same culture. Microcarrier; Neurons; Glial cell; Cell separation; CNS; Immunocytochemistry
Introduction
The classical in vitro method for studying function and differentiation of dissociated cells from the central nervous system (CNS), involves growing bidimensional Correspondence to: Dr. A. Shahar, Section of Electron Microscopy, Dept. of Virology, Israel Institute for Biological Research, P.O. Box 19, Ness-Ziona 70450, Israel. 0168-1656/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
222 monolayer cultures on polylysine coated supports (Yavin and Yavin, 1977; Sensenbrenner, 1977). One way of obtaining a three-dimensional growth pattern of CNS cells is by cultivating them on microcarriers (MCs). In these MC-cultures, cells are propagated on small solid particles suspended in the culture medium. Several types of MCs having different forms and sizes are commercially available (Reuveny, 1983). We have cultured dissociated cells of the CNS on positively charged cylindrical microcarriers on which they grow and differentiate in a tridimensional pattern (Shahar et al., 1983; 1984a). Such MC-cultures are grown either in rotating flasks or in stationary plastic dishes (Shahar and Reuveny, 1987; Shahar, 1990). In both these methods, cells attach to the MCs and form aggregates which usually remain floating in the medium during cultivation. In the rotating system, due to the rotation, all types of cultured cells become firmly attached to the MCs which form aggregates. In stationary cultures, during the first period of cultivation, the cells are not yet tightly interconnected, thus when aggregates sediment near the bottom of the dish, a number of glia and flat cells detach from the aggregates and adhere to the plastic. In this way two separate cultures are obtained: in one culture most of the neuronal components remain attached to the MCs of the floating aggregates and grow in a tridimensional pattern. In the other culture, the non neuronal elements containing meningeal and glial cells, as well as fibroblasts, grow in the form of a monolayer on the plastic of the dish. For the identification of the cell types, glial fibrillary acidic protein (GFAP) was employed. It is widely used as an immunohistochemical marker for astrocytic cells (Cordell et al., 1984). Since G F A P was found to be a major constituent of glial fibers, the expression of G F A P has been interpreted as a marker for astroglial histogenesis a n d / o r differentiation. Neuron specific enolase (NSE) has been introduced as a marker of neuronal histogenesis and differentiation by several authors (Pickel et al., 1976; Schmechel et al., 1978). Kumpulainen and Korhonen (1982) reported that mouse and human oligodendrocytes are Carboanhydrase C (CA-C)-positive, and therefore CA-C-antiserum can be used as a marker for normal oligodendrocytes. In vivo results have shown that in rat brain, the most active synthesis of this enzyme occurred during the myelination period. In glial cell cultures, the level of CA increased in parallel to the development of the oligodendroglia-like cells and the enzyme was only localized in these cells (Delaunoy et al., 1989). In this study, we used scanning electron microscopy (SEM) to determine the tridimensional growth of neurons on MCs, and immunocytochemistry for the characterization of glia cells grown on the plastic.
Material and Methods Microcarriers
We have used cylindrical MCs (DE-53) produced by Whatman, U.K. and marketed as an anion exchange resin for chromatography. These MCs measure
223 80-400/~m in length and 40-50/~m in diameter. They are made of cellulose matrix, charged with tertiary amine groups. For use, they are equilibrated with phosphatebuffered saline (pH 7.4) and autoclaved (121 ° C, 20 min) in batches of 5 g in 100 ml. After being washed twice in nutrient medium. MCs are added to cultures to a final concentration of 3-5 mg per 1 0 6 ceils. Since MCs tend to stick to glass, they should be handled with either plastic or siliconized glass instruments. For further details on the use of MCs, see Reuveny (1983) and Shahar (1990).
MC-cultures Brain and spinal cords were dissected, from Sprague-Dawley rat fetuses at the ages of 16 and 14 d, respectively. Cells were obtained by mechanical dissociation (Yavin and Yavin, 1977) and were suspended with the MCs as follows: For rotating cultures, 8-10 x 1 0 6 cells and 25-30 mg of MCs were suspended in 50 ml Erlenmeyer shake flasks containing 10 ml of nutrient medium. Cultures were rotated at 150 rotations per min (1 inch amplitude). For stationary cultures 1 x 1 0 6 cells and 3 mg of MC cells were plated in 32 mm plastic dishes without substrate coating, in 1.5 ml of nutrient medium per dish. The culture medium for brain cells consisted of 20% fetal calf serum and 77% Eagle's basal medium. The spinal cord culture medium contained 10% of each fetal calf serum and horse serum and 77% of MEM-Eagle's. Both media were supplemented with 1% of 2 mM L-glutamine, 1% glucose (from a 50% solution to provide a final concentration of 600 mg) and of 1% 16/~g ml -a gentamicin solution. The sera were heat inactivated at 56 o C for 25 min. Stationary cultures were incubated in a CO2 incubator and their medium was changed twice a week. On the second, and sometimes also on the third feeding of the stationary cultures, the aggregates were carefully transferred to new uncoated plastic dishes, using a curved Belco pipette. The remaining dishes, with the attached cells, were cultured for a few more days prior to fixation. The rotating cultures were pregassed with 5% CO 2 in air and closed with rubber stoppers. Only 50% of the medium was replaced twice a week.
Processing of aggregates for SEM Aggregates were collected in conic tubes and fixed in 2.5% glutaraldehyde, postfixed in 1% osmium tetroxide and dehydrated in increments of ethanol. Critical point drying was made in a Polaron apparatus using liquid CO 2. Aggregates were then placed on double scotch tape and vacuum coated with 200 A gold-palladium (Polaron sputtering unit). Observation was made in a JSM 35C at 25 KV.
Immunocytochemistry G F A P for astroglial cells was used in the alkaline phosphatase-anti alkaline phosphatase (APAAP) technique according to Cordell et al. (1984). Dilution of GFAP-antibodies was 1 : 250 (Dakopatts, Denmark). NSE as a neuron specific marker and CA-C as a marker of oligos were used according to the techniques described by Nakagawa et al. (1986). The dilution of anti-neuron-specific enolase was 1:1000 (Dakopatts, Denmark) and that of anticarbonic anhydrase C was 1 : 50 (Calbiochem-Behring).
Fig. 1. SEM of MC-aggregates 15 d in vitro (DIV). A. Flat cells exhibiting numerous microvilli in spinal cord cultures. B. Perikarya and a network of interconnecting fibers of brain cells.
225 Results
In rotating CNS cultures, mainly of spinal cord, in which meninges had not been stripped off, most of the aggregates during the entire period of cultivation were composed of a mixed cell population. The cells were composed, in certain areas, of round or fusiform perikarya, which formed a densely packed network of nerve fibers among them (Fig. 1B). Other areas were covered with a sheath of adjacent fibroblasts and meningeal ceils exhibiting numerous microvilli. With time, the non-neuronal elements divided and ensnared most of the surface of the aggregates (Fig. 1A). On the other hand, in stationary cultures, due to detachment of a large number of glial cells, meningeal elements and fibroblasts, the aggregates were composed mainly of nerve cells, which adhered directly to the MCs and sprouted to form a well organized fiber network (Figs. 2, 3). In some aggregates, mainly of brain cultures, several parts were covered by ependymal cells which remained attached to the MCs and formed epithelial like laminae (Fig. 4A). The luminal surface of the ependymal cells was convex and covered with numerous microvilli, many of the cells exhibited groups of cilia (Fig. 4B). Aggregates which were formed in spinal cord MC-cultures contained, in addition to neuronal cells, also elongated myotubes which were arranged adjacent and parallel to each other (Shahar et al., 1986). A few aggregates in stationary cultures might attach to the plastic of the dish and act as explants with nerve fibers sprouting over a layer of migrating non neuronal elements (Fig. 5A). In spinal cord MC-cultures a few contracting multinucleated myotubes can be observed in the migration zone. However, beside these single attached aggregates with their migration cell zones, the entire plastic dish, in stationary MC-cultures, was mainly covered by a confluent monolayer made of detached cells. The predominating glial cell type in the confluent monolayer culture was considered to be astroglial because the bodies of these cells, and their processes were intensively stained with labelled antibodies against GFAP. These cells were of the average size of 1 5 - 2 0 / , m and were either flat or spindle-shaped with a few processes. Their unstained central oval nuclei contained usually one or two nucleoli (Fig. 5B). The other glial cell type which can be recognized in the monolayer cultures is the oligodendrocyte. These cells appear in considerable numbers and stain positively with anti-carboanhydrase (Fig. 6A). They are small cells with a central prominent nucleus surrounded by a thin cytoplasmic area. Cytoplasmic processes are not always recognized. In the monolayer, there were no NSE positively stained neurons (Fig. 6B). However, in the vicinity of attached aggregates, sometimes single neurons could be observed. Cells in the background which do not stain with immunocytochemical dyes are flat meningeal cells, fibroblasts, as well as round refractile macrophages. Only during the first week of cultivation (prior to the first and perhaps the second transfer of the neuronal aggregates), glial elements and flat cells detached to form a monolayer. During the following weeks in culture, usually no more non-neuronal elements detached.
226
227
Fig. 3. SEM showing two neurons and their processes attached to MCs in the aggregate. Note some flat cells remaining in A. Brain cultures, 20 DIV.
Fig. 2. Intensive network of neuronal fibers and a few rounded and flattened perikarya which remain in the aggregates. Note in A and B the peculiar growth of nerve fibers directly attached to the MCs--unlike in monolayer cultures in which they usually sprout above a layer of dividing non neuronal elements. In C elongated nerve fibers ramify to interconnect with cells which are attached to distant MCs within the same aggregates. SEM, brain cells, 15 DIV.
228
229
Fig. 5. Light microscopy showing in: A. Cell migrating from an aggregate attached to the plastic. B. Astroglial cells stained with labelled antibodies against GFAP, 9 DIV. Original magnifications: A, x 150; B, x 325.
Fig. 4. SEM of ependymal cells from a 14 d brain culture. Cells are organized on the MCs in a large epithelial layer which is probably formed following division of the ependymal cells. The ceils exhibit groups of cilia and microvilli.
230
Fig. 6. Light microscopy of glial cells detached from stationary aggregates after 9 DIV. A, Oligos stained with anticarbonicanhydrase; B, NSE stain showing absence of neurons in the monolayer. Original magnification x 150.
231
Discussion
We have previously reported on the use of MC-cultures, either in stationary or rotating system, as a tool for the study of CNS cell differentiation in vitro. Dissociated brain cells, when grown on MCs, develop synaptic units and become myelinated (Shahar et al., 1983). In spinal cord MC-cultures, in addition to nerve cell differentiation, there is also growth of muscle fibers which mature to form nerve-muscle interconnections (Shahar et al., 1984, 1987). The aggregates in the stationary nerve-muscle cultures were not transferred during the entire cultivation period and were composed of both neuronal and non-neuronal elements. A similar growth pattern of heterogenous cell population of floating aggregates has been described by others in aggregating CNS cultures maintained in stationary and rotating systems (Seeds et al., 1977; Bjerkvig et al., 1986; Bjerkvig, 1986). Here we describe a different approach for cultivation of dissociated CNS on MCs so as to separate different cellular components into two culture types: 1. The neuronal elements, ependyma (and in spinal cord cultures also muscle fibers) remain attached to the MCs forming aggregates in which they establish in a tridimensional pattern. 2. Following the transfer of the neuronal aggregates, the remaining detached glial, meningeal and other non-neuronal elements develop into a monolayer on the plastic dish. Immunocytochemical stains indicate that the predominant glial cell type in the monolayer are astrocytes, though there is also a considerable number of oligos present. This separation of cellular components by the stationary MC-culture technique has several advantages: In the first place it allows to study the specific effects induced by chemical, physical and biological agents acting on either the nerve and the nerve-muscle cell population or on the glia and meningeal elements. Furthermore, on the metabolic-biochemical level, the secretion of factors, neurotransmitters, metabolites, etc., by either of the cellular components can be analysed and evaluated. Finally, the neuronal or nerve-muscle MC entities are easily transferred to other monolayer cultures of different tissues, or can be implanted into living, injured or dystrophic, tissues.
References Bjerkvig, R. (1986) Reaggregation of fetal rat brain cells in stationary culture system II: Ultrastructural characterization. In Vitro Cell Dev. Biol. 22, 193-200. Bjerkvig, R., Steinsvag, S.L. and Laerum, O.D. (1986) Reaggregation of fetal rat brain cells in a stationary culture system I: methodology and cell identification. In Vitro Cell Dev. Biol. 22, 180-192. Cordell, J.L., Falini, B., Erber, W.N., Ghosh, A.K., Abdulaziz, Z., MacDonald, S., Pulford, K.A.F., Stein, H. and Mason, D.Y. (1984) Immunoenzymatic labelling of monoclonal antibodies using immune complexes of alkaline phosphatase and monoclonal antialkaline phosphatase (APAAP complexes). J. Histochem. Cytochem. 32, 219-229. Delaunoy, J.P., Hog, F., Devilliers, G., Bansart, M., Mandel, P. and Sensenbrenner, M. (1980) Developmental changes and localization of carbonic anhydrase in cerebral hemispheres of the rat and rat glial cell cultures. Cell. Mol. Biol. 26, 235-240. Kumpulainen, T. and Korhonen, L.K. (1982) Immunohistochemical localization of carbonic anhydrase isoenzyme C in the central and peripheral nervous system of the mouse. J. Histochem. Cytochem. 30, 283-292.
232 Nakagawa, Y., Perentes, E. and Rubinstein, J.L. (1986) Immunohistochemical characterization of ohgodendrogliomas: an analysis of multiple markers. Acta Neuropathol. 72, 15-22. Pickel, V.M., Reis, D.J., Marangos, P.J. and Zomzely-Neurath, C. (1976) Immunocytochemical localization of nervous system specific protein (NSP-R) in rat brain. Brain Res. 105, 184-187. Reuveny, S. (1983) Microcarriers for culturing mammalian cells and their application. In: A. Mizrahi and A.L. Van Wezel (Eds.), Advances in Biotechnological Processes, Alan R. Liss Inc., New York, Vol. 2, pp. 1-32. Schmechel, D., Marangos, P.J., Zis, A.P., Brightman, M. and Goodwin, F.K. (1978) Brain enolases as specific markers of neuronal and glial cells. Science 199, 313-315. Seeds, N.W., Ramirez, G. and Marko, M.J. (1977) Aggregates cultures: a model for studies of brain development. In: R.T. Acton and J.D. Lynn (Eds.), Cell Culture and its application, Academic Press, New York, pp. 23-37. Sensenbrenner, M. (1977) Dissociated brain cells in primary cultures. In: S. Federoff and L. Hertz (Eds.), Tissue and Organ Cultures in Neurobiology, Academic Press, New York, pp. 191-213. Shahar, A., Reuveny, S., Amir, A., Kotler, M. and Mizrahi, A. (1983) Synaptogenesis and myelination in dissociated cerebral microcarrier cell culture. J. Neurosci. Res. 9, 339-348. Shahar, A., Amir, A., Reuveny, S., Silberstein, L. and Mizrahi, A. (1984a) Neuronal cultures in microcarriers: Dissociated spinal cord cells. Develop. Biol. Standard 55, 25-30 S. Karger, Basel. Shahar, A., Reuveny, S., Mizrahi, A. and Shainberg, A. (1984b) Differentiation of dissociated embryonic central nervous system cells and of myoblasts cultured on microcarriers. J. Acad. Med. Torino CXL VII, 33-38. Shahar, A. and Reuveny, S. (1987) Nerve and muscle cells on microcarriers in culture. In: A. Fiechter (Ed.), Advances in Biochemical Engineering, Vol. 34. Shahar, A. (1990) Cultivation of nerve and muscle cells on microcarriers. Methods Neurosci. 2, 195-209. Shahar, A., Bidder, M., David, Y., Amir, A. and Shainberg, A. (1987) Autologous neuronal and muscle co-cultures: a model for neurotoxicology. In: A. Shahar and A.M. Goldberg (Eds.), Model Systems in Neurotoxicology. Alternative Approach to Animal Testing, Alan R. Liss Inc., New York, pp. 45-58. Yavin, Z. and Yavin, E. (1977) Synaptogenesis and myelinogenesis in dissociated cerebral cells from rat embryo on polylysine coated surfaces. Exp. Brain Res. 29, 137-147. Yavin, E. and Yavin, Z. (1974) Attachment and culture of dissociated cells from rat embryo cerebral hemispheres on polylysine-coated surface. J. Cell Biol. 62, 540-546.
