- Email: [email protected]
Differentiation of purified astrocytes in a chemically defined medium
DevelopmentalBrain Research, 9 (1983) 337-345 Elsevier
337
Differentiation of Purified Astrocytes in a Chemically Defined Medium RICHARD S. MORRISONL4 and JEAN DE VELLIS 1-5
Departments of tAnatomy and 2Psychiatry, School of Medicine, 3MentalRetardation Research Center, 4Laboratoryof Biochemical and Environmental Sciences, and 5Brain ResearchInstitute, Universityof California, Los Angeles, CA 90024(U.S.A.) (Accepted February 22nd, 1983)
Key words: astrocytes - - growth factors - - differentiation - - glial fibrillary acidic protein - - CNS - - cell culture
Homogeneous cultures of astrocytes and oligodendrocytes provide an excellent model system for studying the regulation of glial structure and function. Recently, a chemically defined (CD) medium was developed for purified cultures of astrocytes, thus eliminating the requirement for serum and providing a controlled system for the study of astroglial properties. Due to the widespread use of astrocyte cultures and the potential benefits to be gained from using a defined medium, astrocyte cultures raised in CD medium were analyzed for purity as well as morphological and biochemical properties. Purity was assessed using immunocytochemical staining for glial fibrillary acidic protein (GFAP) and fibronectin. Astrocytes raised in CD medium are 95% pure using the expression of GFAP as a criterion. Fewer than 1% of the cells in CD medium stained positive for fibronectin eliminating the possibilitythat CD medium is selective for meningeal or endothelial cells. Astrocytes raised in CD medium exhibit a striking degree of morphological differentiation as seen in scanning electron micrographs. They also exhibit a high degree of biochemical differentiation illustrated by increases in the specific activity of S-100 protein and the induction of glutamine synthetase by glucocorticoids. A defined medium that supports the proliferation of rat astrocytes and enhances numerous morphological and biochemical properties should greatly facilitate the study of factors controlling glial proliferation and differentation.
INTRODUCTION To elucidate the mechanisms governing the develo p m e n t of the nervous system it is necessary to identify factors which regulate the proliferation and differentiation of neural cells. T h e search for substances which are capable of influencing neural d e v e l o p m e n t has met with some success in investigations of both the PNS 6,13,18,24--26,32 and CNS 9,23,29. These advances in n e u r o b i o l o g y have been greatly aided, in part, by the d e v e l o p m e n t of in vitro cell culture systems. This is especially true for the PNS from which it is possible to p r o d u c e highly purified cultures of neurons and supporting cells, Progress in the analysis of central neurogenesis has been h i n d e r e d , however, due to difficulties in p r e p a r i n g purified cultures of the various neural cell types. A l t h o u g h a m e t h o d is still not available for p r o d u c i n g h o m o g e n o u s cultures of CNS neurons, it is now possible to g e n e r a t e highly purified cultures of astrocytes and o l i g o d e n d r o c y t e s 16. Previous studies on astroglial proliferation and differentiation have utilized fetal calf serum as a component in the culture mediumSA2,16,20,28.30. The unde0165-3806/83/$03.00 © 1983 Elsevier Science Publishers B.V.
fined nature of serum and the variability of many of its c o m p o n e n t s from one batch to a n o t h e r was a complicating factor in these studies, reducing the control one has over environmental factors in cell culture. In a recent communication 17 we r e p o r t e d that the requirement for serum could be circumvented by growing purified astrocytes in a chemically defined (CD) medium. W e now r e p o r t that astrocytes raised in C D m e d i u m exhibit e n h a n c e d morphological and biochemical differentiation. MATERIALS AND METHODS
Cell culture Purified cultures of rat astrocytes were p r e p a r e d from p r i m a r y cerebral cultures by the m e t h o d of McCarthy and de Vellis 16. Meningeal cell cultures were p r e p a r e d by removing the meninges from the cortices of 0 to 1-day-old rat pups, and dissociating the tissue with trypsin for 30 min. Trypsinization was stopped by adding culture m e d i a s u p p l e m e n t e d with 10% FCS. The cell suspension was centrifuged for 5 min at 800 rpm in an International Clinical Centrifuge. The
338 supernatant was discarded and the meningeal cells were resuspended in fresh serum-supplemented medium for plating. Both astroglial and meningeal primary cell cultures were initially plated in Falcon plastic tissue culture flasks (75 cm2), and maintained in Hams F-12 medium/Dulbecco-Vogt modification of Eagles medium (1:1 v/v), with 1.2 g/l of NaHCO 3, 15 mM hepes buffer (serum free (SF) medium) and 10% (v/v) FCS. Purified cultures of astrocytes and meninges were trypsinized with a 1% Enzar-T (40x stock concentrate) trypsin concentrate and 0.1 mM EDTA in Hank's balanced salt solution (Ca 2+- and Mg2+-free) and then diluted 1:4 in medium containing 10% FCS. The cells were counted in a Royco 927T cell tissue counter and plated out at the appropriate densities. Cells were plated directly on plastic unless stated otherwise. All cell cultures were maintained in a humidified atmosphere of 5% CO2-95% air at 37 °C.
CD medium This consisted of the SF medium plus 50 mM hydrocortisone, 100 pM putrescine, prostaglandin F2a (PGF2,, 500 ng/ml), insulin (50 ~g/ml), and Fibroblast Growth Factor (FGF, 100 ng/ml).
dilution), washed extensively, and then stained with fluorescein isothiocyanate (FITC)-conjugated swine anti-rabbit IgG (1:50 dilution; 30 min). After additional washing the coverslips were rinsed in distilled water and mounted on glass slides with glycerol/phosphate-buffered saline, 1:9 (v/v). Cells were observed through a Zeiss fluorescence microscope equipped with FITC filters and mercury-vapor epi-illumination.
Scanning electron microscopy (SEM) Cultures were rinsed twice with physiological saline (0.85%), and then fixed with 4% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4, 0.05% calcium chloride added) for 2 h at room temperature. Following the rinse with buffer, cultures were postfixed with 1% osmium tetroxide in the same cacodylate buffer for 2 h. Cultures were then dehydrated in graded ethyl alcohols. After dehydration, the specimens were critically point dried from CO2. They were coated for 2 min using 15 s intervals with gold palladium in a Hummer I coater. Cells were observed with an ETEC Autoscan microscope.
Enzyme assays Glial fibrillary acidic protein (GFAP) and fibronectin immunofluorescence Astroglial and meningeal cell cultures were grown on 15 mm glass coverslips contained in microwell plates (Falcon). Astrocytes were seeded at 1 x 104 per well (2.1 cm2) in serum-supplemented medium for 18 h, rinsed 3× with SF medium and then switched to CD medium. Meningeal cells were seeded at 2 × 104 per well (2.1 cm 2) in serum-supplemented medium for 48 h, rinsed 3 times with SF medium and then switched to CD medium. Control cultures were maintained in serum-supplemented medium for the duration of the experiment. After 7 or 14 days in CD medium (7 days for meningeal cell cultures) cell cultures were washed with phosphate-buffered saline (pH 7.4) and fixed at room temperature for 15 rain in 2% (w/v) paraformaldehyde (0.1 M sodium cacodylate buffer/0.05% calcium chloride, pH 7.4). Cultures used for the study of GFAP were fixed for an additional 5 min with acetone. The cells were incubated for 30 rain with anti-GFAP rabbit antiserum (1:200 dilution), or anti-fibronectin rabbit antiserum (1:160
The methods used to assay glycerol phosphate dehydrogenase (GPDH) and glutamine synthetase (GS) have been previously described2,22. One unit of enzyme activity is defined as the transformation of one nmole of substrate per min at 30 °C for GPDH and 37 °C for GS. Specific activity is expressed as units of enzyme activity per mg of total cell protein. All assays were performed in triplicate.
S-IO0 protein Values of S-100 protein were determined as described by Zuckerman et al. 33 by the micro-complement fixation of LevineU.
Protein assay Protein content was determined by the method of Lowry et al.~4 using crystallized bovine serum albumin as a standard.
Materials Materials were obtained from the following sources: crystalline bovine insulin, PGF2a, and putrescine
339 dihydrochloride from Sigma; FGF from Collaborative Research; hydrocortisone acetate from Calbiochem-Behring; Ham's F-12, DMEM, and Hank's Balanced Salt Solution (Mg 2÷- and Ca2+-free) from Gibco; fetal calf serum from Irvine Scientific; trypsin from Reheis; fluorescein conjugated swine anti-rabbit IgG from Bio-Rad Laboratories; guinea pig complement and sheep red blood cells from M. A. Bioproducts; hemolysin from BBL; division of Becton, Dickinson; and L-[U-14C]glutamic acid (285 mCi/~mol) from Amersham. The following gifts are gratefully acknowledged: anti-GFAP rabbit antiserdm supplied by Dr. Larry F. Eng (Stanford University and V.A. Hospital, Palo Alto), and S-100, anti-S-100 rabbit antiserum and antifibronectin rabbit antiserum supplied by Dr. Harvey R. Herschman (UCLA).
RESULTS Cultures of purified astrocytes raised in CD medium were evaluated with respect to their purity and state of differentiation. The differences between astrocyte cultures raised in the presence of fetal calf serum or in defined medium have not been previously reported. To address this question, we grew purified astrocytes in parallel cultures under these two conditions. Astrocytes grown in serum-supplemented medium (10% FCS) for one or two weeks appear flat and morphologically undifferentiated (Fig. 1A). These cells possess GFAP as demonstrated by indirect immunofluorescence (Fig. 1B). An analysis of randomly chosen fields on several tissue culture dishes showed that 90% of the cells specifically stained for GFAP
Fig. 1. GFAP immunofluorescence of astrocyte cultures grown in serum-supplemented medium (10%). The same field from a representative culture was visualized by phase-contrast microscopy(A) and indirect immunofluorescencefor GFAP (B). Substitution of
GFAP-antisera by normal rabbit serum IgG displayed no visible fluorescence. All other contr*ols,including blocking the antiserum with purified GFAP and reagent controls, were also negative (x 520).
340 TABLE I Temporal expression of'cell markers
Imm~mocytochemistrywas performed as described in Materials and Methods. The data represent the percentage of cells expressing the specific marker in relation to the total cell population. The value of S-100 protein at 2 weeks in culture was significantly greater in CD medium than in serum-supplement medium (P ~<0.001). 1 Week
GFAP Fibronectin S-100 ~g/mg protein)
2 Weeks
Serum
CDM
Serum
CDM
>90.00% < 5.00% 0.40+0.06
>95.00% < 1.00% 0.25+0.06
>90.00% < 5.00% 1.20+0.09
>95.00% < 1.00% 1.42+0.07
(Table I). Astrocytes expressing G F A P exhibited fluorescence in the soma and processes. Nuclei were never observed to fluoresce. The cells raised in CD medium were 95% astrocytic as judged by their expression of G F A P (Table I). The cellular distribution of G F A P was the same as that observed for cells in serum-supplemented medi-
um, but the immunofluorescence clearly accentuated the enhanced formation of processes observed in cells grown in CD medium (Fig. 2B). In addition to the greater degree of cellular processes, G F A P immunofluorescence was much more intense in cells grown in CD medium. Astrocyte cultures raised in CD medium were also
Fig. 2. GFAP immunofluorescence of astrocyte cultures grown in CD medium. The same field from a representative culture was visualized by phase-contrast microscopy (A) and indirect immunofluorescence for GFAP (B). For controls, see Fig. 1 (x 520).
341
Fig. 3. Fibronectin immunofluorescence of meningeal cell cultures. Meningeal cells were plated in serum-supplemented medium (10%, 48 h) and converted to CD medium for 7 days. The same field from a representative area was visualized by phase-contrast microscopy (A) and indirect immunofluorescence for fibronectin (B) (x 520).
examined for the appearance of fibronectin. Fibronectin is considered a reliable marker of meningeal and endothelial cell populations27, 30. A random sampling of different culture fields did not show appreciable staining for fibronectin utilizing indirect immunofluorescence (data not shown). The percentage of fibronectin-positive cells was less than 1% (Table I). In contrast, cultures raised in serum-supplemented medium exhibited occasional loci of fibronectin-positive cells (data not shown). The percentage of cells expressing fibronectin in these cultures never amounted to more than 5% of the total cell population. Although the difference between the number of fibronectin-positive cells found in serum-supplemented and CD medium is not great under normal conditions, it can be accentuated. The meninges are normally removed when primary cultures of cerebrum are established. If this step is not carefully performed the percentage of fibronectin-positive cells is drastically increased. In cultures prepared in this manner they can represent as much as 36% and 20% of the total cell population in serum-supplemented and defined medium cultures, respectively. The near complete absence of fibronectin in cul-
tures raised in CD medium cannot be attributed solely to the removal of serum. Purified meningeal cultures were established as controls. Meningeal cells plated and raised in serum-supplemented medium display a pericellular matrix of fibronectin (data not shown). If the cells are plated in serum-supplemented medium for 2 days, and then shifted to the defined medium for 7 days the fibronectin matrix is still present (Fig. 3). The shift to defined medium did not visibly alter the structure of the fibronectin matrix. Astrocytes maintained in serum-supplemented medium are flat and possess few processes (Fig. 4A,C). In addition, their surfaces are smooth and devoid of any distinguishable features. These characteristics also apply to astrocyte cultures maintained in SF medium after the initial 18 h serum-preincubation (data not shown). Astrocytes maintained in CD medium, however, exhibited dramatic changes in cell morphology. The majority of cells exhibit small ovoid somas and long branching processes (Fig. 4B,D). A small population of cells, however, possessing the flat, polugonal shaped morphology observed in serum-supplemented cultures remains in cultures grown in CD medium.
342
C
D
Fig. 4. Scanning electron micrographs of astrocyte cultures raised in serum-supplemented medium (A) and CD medium (B) (x 1680). Astrocytes were plated directly into serum-supplemented or CD medium and maintained for 4 days before processing. Sections of Araldite-embedded cultures (cut perpendicular to plating surface and stained with Toluidine blue) are shown for cells grown in serumsupplemented (C) and CD medium (D). Cells in CD medium display numerous filopodia (B, arrow). Bar equals 5 #m.
The somas of process-bearing cells possessed many convoluted folds and the processes exhibited evenly spaced arrays of filopodia (Fig. 4B). The tiMpodia were not present where processes appeared to be forming in the transition from a flattened morphology to the process-bearing morphology. These flattened segments clearly resemble the morphology of cells grown in serum-supplemented medium. The most notable difference between processbearing cells was the variable number of processes found issuing forth from the cell body. Astrocytes grown in CD medium exhibit not only a striking degree of morphological differentiation but
also a high degree of biochemical differentiation. Aside from expressing G F A P several other properties appear to be enhanced in these cells (Table I). S-100 protein, a defined glial marker7 increased 3- and 6-fold from week 1 to week 2 in serum-supplemented and CD medium, respectively. S-100 protein is increased from 1.20 ~g/mg protein in serum-supplemented cultures to 1.42/~g/mg protein in CD medium at week two. S-100 protein was not detected, however, in control cultures of rat oligodendrocytes, bovine endothelial cells and rat skin fibroblasts (data not shown). In the CNS, the enzyme glutamine synthetase (GS) is also restricted in location to glial
343
~_ t4-A' +Serum
most commonly employed. Unfortunately, the inclusion of serum has several disadvantages. Serum contains many undefined components, some of which may actually inhibit cell growth or interact with substances being studied. Furthermore, defined serum components commonly exhibit variation from batch to batch thereby influencing experimental reproduc-
B CDM
8
i
GS
fil
GPDH
n n= GS
GPDH
Fig. 5. Enzyme assays were performed as described in Materials and Methods. To measure the induction of GS, astrocytes were plated at 105 cells/35 mm dish in serum-supplemented medium. Eighteen hours later, half of the dishes were converted to CD medium minus hydrocortisone. These conditions were maintained for 3 days at which time cultures in both serum-supplemented and CD medium were induced for 48 h, with 1 pm hydrocortisone. All the media were replenished after the first 24 h incubation. Hydrocortisone induction of glutamine synthetase in CD medium was significantlygreater than the induction observed in serum-supplemented medium (P < 0.01). There was no significant induction of GPDH by hydrocortisone under either conditions.
cells ~5 and has been shown to be regulated by glucocorticoids 28. In CD medium, hydrocortisone (1/~M) induced GS activity approximately 4-fold over control cultures, compared to only a 1.5-fold induction in serum-supplemented medium (Fig. 5). Basal levels of GS activity were substantially increased in response to the presence of 50 nM hydrocortisone which was included as part of the growth medium (data not shown). Glycerol phosphate dehydrogenase (GPDH) is another glial enzyme regulated by glucocorticoids 2. The induction of G P D H in neural tissue appears to be restricted to oligodendrocytes ~0. In accordance with this observation we observed that although basal levels of G P D H are detectable, G P D H is not inducible in astrocyte cultures maintained in either CD or serum-supplemented medium (Fig. 5). DISCUSSION Dissociated cultures of neural tissue provide an excellent model system for studying neural structure and function. Cells in culture, however, have traditionally required an organ extract of one type or another as a supplement for plating and proliferation. Serum, usually from fetal animals, is the supplement
ibility. In order to more effectively study neural cells in culture it would be useful to define their nutritional and hormonal needs. A significant start can be made by removing serum from the growth medium. We approached this goal by developing a CD medium for purified cultures of astrocytes. The kinetics of astrocyte proliferation in the CD medium have been previously described 17. Astrocyte cultures raised in CD medium possess more GFAP-positive cells than sister cultures raised in serum-supplemented medium. The difference is not usually significant, but the trend has been observed in all of our experiments. More striking, however, is the difference in the intensity of staining for GFAP. GFAP immunofluorescence is much more intense in astrocyte cultures grown in CD medium. Although immunofluorescence is not a quantitative method it suggested that astrocytes raised in CD medium might possess greater quantities of GFAP. Recent data obtained by R I A demonstrates that astrocytes raised in CD medium contain 2-fold more GFAP (spec. act.) than control cultures raised in serum-supplemented medium (Morrison, Eng and de Vellis, unpublished observations). Astrocyte cultures rarely exhibited contamination by meningeal or endothelial cells as judged by the paucity of fibronectin immunofluorescence. Nevertheless, when fibronectin-positive cells were present they were much more common in cultures grown in serum-supplemented medium. This increase may reflect the incorporation of serum-fibronectin into a pericellular matrix already surrounding astrocytes. Several workers have demonstrated that exogenous fibronectin may be incorporated into the extracellular matrices of cells in culturea,L A more likely explanation, however, is that meningeal and endothelial cells require one or more serum components for survival and proliferation which are lacking in the defined medium. Several other reports on the use of defined medium mention a reduction in contamination by fibroblastic elements1.31. The reduction of fibro-
344 nectin immunofluorescence in our cultures is not due to the removal of serum. Meningeal cells maintained for 7 days in defined medium still display a fibronectin matrix that is not visibly altered from the matrix observed in serum-supplemented cultures. The most salient difference between astrocytes raised in serum-supplemented or C D medium is the dramatic change in morphology. In serum-supplemented medium astrocytes presented a fiat, undifferentiated, epithelioid morphology. In contrast, astrocytes raised in CD medium gradually extended long processes with evenly spaced arrays of filopodia, a morphology reminiscent of the mature astrocyte in vivo 19. A detailed study of the effects of the individual supplements on morphology has not been done. The interplay between all or several of the factors may be necessary for this morphological maturation. It is clearly not the result of simply removing serum, which has no long-term effect on the morphology of these cells. It is interesting to note that soluble extracts from brain can also stimulate the morphological maturation of glial cellsl2, 20. These substances may be growth factors similar to the components found in the defined medium. In support of this notion it has been reported that these morphological maturation factors also possess mitogenic activityS, 20. Astrocytes raised in CI~ medium express biochemical properties that are either maintained or enhanced with respect to cultures raised in the presence of fetal calf serum. S-100 protein for example, is significantly elevated (approximately 18%) in astrocyte REFERENCES 1 Barnes, D. and Sato, G., Methods for growth of cultured cells in serum-free medium, Analyt. Biochem., 102 (1980) 255-270. 2 Breen, G. A. M. and de Vellis, J., Regulation of glycerol phosphate dehydrogenase by hydrocortisone in dissociated rat cerebral cell cultures, Develop. Biol., 41 (1974) 255-266. 3 Gysin, R., Moore, B. W., Proffitt, R. T., Devel, T. F., Caldwell, K. and Glaser, L., Regulation of the synthesis of S-100 protein in rat glial cells, J. biol. Chem., 255 (1980) 1515-1519. 4 Hayman, E. G. and Ruoslahti, E., Distribution of fetal bovine serum fibronectin and endogenous rat cell fibronectin in extracellular matrix, J. Cell Biol., 83 (1979) 255-259. 5 Hayman, E. G., Engvall, E. and Ruoslahti, E., Concomitant loss of cell surface fibronectin and laminin from transformed rat kidney cells, J. Cell Biol., 88 (1981) 352-357. 6 Hendry, I. A., Cell division in the developing sympathetic nervous system, J. Neurocytol., 6 (1977) 299-309.
cultures raised in CD medium. The regulation of S-100 protein is not understood, but the data that are available are in part contradictory3,2k While it is agreed that S-100 protein is regulated by homotypic cellular contacts in C6 glioma cells, the effect(s) of peripheral agents such as hormones and growth factors is controversial. The development of a CD medium that promotes attachment, proliferation and differentiation of astrocytes marks a significant advance in our ability to study astrocytes in vitro. The defined conditions should facilitate the study of the molecular mechanisms controlling astroglial proliferation and differentiation, which will hopefully allow us to rectify controversies like the one surrounding the regulation of S-100 protein. ACKNOWLEDGEMENTS This work was supported by the Department of Energy Contract DE-AM03-76-SF00012, Cellular and Molecular Biology U.S. Public Health Service Training Grant GM07185, and U.S. Public Health Service Grants HD-05615 and AG-01754. We wish to thank Dr. Alaric T. A r e n a n d e r for critically reading the manuscript and Dr. Peter Shintaku for his expert assistance with the scanning electron microscopy and illustrations. We also thank Joyce Adler and Mary A n n Cook for their help in preparing the manuscript.
7 Hyden, H. and McEwen, B. S., A glial protein specific for the nervous system, Proc. nat. Acad. Sci. U.S.A., 55 (1966) 354-358. 8 Kato, T., Fukui, Y., Turriff, D. E., Nakagawa, S., Lim, R.. Arnason, B. G. W. and Tanaka, Glial maturation factor in bovine brain: partial purification and physicochemical characterization, Brain Res., 212 (1981) 393--402. 9 Landmesser, L. and Pilar, G., Interactions between neurons and their targets during in vivo synaptogenesis, Fed. Proc., 37 (1978) 2016-2022. 10 Leveille, P. J., McGinnis, J. F., Maxwell, D. S. and de Vellis, J., Immunocytochemical localization of glycerol-3phosphate dehydrogenase in rat oligodendrocytes, Brain Res., 196 (1980) 287-305. 11 Levine, L., Microcomplement fixation. In D. M. Weir (Ed.), Handbook of Experimental lmmunochemistry, Blackwell Scientific Publications, Oxford, 1967, pp. 707-719. 12 Lim, R., Glia maturation factor, Current Topics in Develop. Biol., 16 (1980) 305-322. 13 Lofberg, J., Ahlfors, K. and Fallstrom, C., Neural crest cell
345
14
15
16
17
18
19
20
21
22
23
24
migration in relation to extracellular matrix organization in the embryonic Axolotl trunk, Develop. Biol., 75 (1980) 148-167. Lowry, O. H., Rosebrough, W. F., Farr, A. C. and Randall, R. V., Protein measurement with the folin phenol reagent, J. biol. Chem., 193 (1951)265-275. Martinez-Hernandez, A., Bell, K. P. and Norenberg, M. D., Glutamine synthetase: glial localization in brain, Science, 195 (1356-1358). McCarthy, K. D. and de Vellis, J., Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue, J. Cell Biol., 85 (1980) 890-902. Morrison, R. S. and de Vellis, J., Growth of purified astrocytes in a chemically defined medium, Proc. nat. Acad. Sci. U.S.A., 78 (1981) 7205-7209. Patterson, P. H., Environmental determination of autonomic neurotransmitter functions, Ann. Rev. Neurosci., 1 (1978) 1-17. Peters, A., Palay, S. L.and Webster, H. def., The Fine Structure of the Nervous System: The Neurons and Supporting Cells, W. B. Saunders, Philadelphia, 1976. Pettman, B., Sensenbrenner, M. and Labourdette, G., Isolation of a glial maturation factor from beef brain, FEBS Lett., 118 (1980) 195-199. Pfeiffer, S. E., Herschman, H. R., Lightbody, J. and Sato, G., Synthesis by a clonal line of rat glial cells of a protein unique to the nervous system, J. cell. Physiol., 75 (1970) 329-339. Pishak, M. R. and Phillips, A. T., A modified radioisotopic assay for measuring glutamine synthetase activity in tissue extracts, Analyt. Biochem., 94 (1979) 82-88. Roger, L. J. and Fellows, R. E., Evidence for thyroxinegrowth hormone interaction during brain development, Nature (Lond.), 282 (1979) 414--415. Salzer, J. L. and Bunge, R. P., Studies of Schwann prolifer-
25
26
27
28
29
30
31
32
33
ation during development, Wallerian degeneration, and direct injury, J. Cell Biol., 84 (1980) 739-752. Salzer, J. L., Williams, A. K., Glaser, L. and Bunge, R. P., Studies of Schwann cell proliferation. II, Characterization of the stimulation and specificity of the response to a neurite membrane fraction, J. Cell Biol., 84 (1980) 735-766. Salzer, J. L., Bunge, R. P. and Glaser, L., Studies of Schwann cell proliferation. III. Evidence for the surface localization of the neurite mitogen, J. Cell Biol., 84 (1980) 767-768. Schachner, M. G., Schoonmaker, G. and Hynes, R. O., Cellular and subcellular localization of LETS protein in the nervous system, Brain Res., 158 (1978) 149-158. Schousboe, A. G., Svenneby, G. and Hertz, L., Uptake and metabolism of glutamate in astrocytes cultures from dissociated mouse brain hemispheres, J. Neurochem., 29 (1977) 999-1005. Smith, R. M., Patel, A. J., Kingsbury, A. E., Hunt, A. and Balazs, R., Effects of thyroid state on brain development: beta-adrenergic receptors and 5'-nucleotidase activity, Brain Res., 198 (1980) 375-387. Steig, P. E., Kimelberg, H. K., Mazurkiewicz, J. E. and Banker, G. A., Distribution of glial fibrillary acidic protein and fibronetin in primary astroglial cultures from rat brain, Brain Res., 199 (1980) 493-500. Taub, M. and Sato, G., Growth of functional primary cultures of kidney epithelial cells in defined medium, J. Cell Physiol., 105 (1980) 369-378. Varon, S. S. and Bunge, R. P., Trophic mechanisms in the peripheral nervous system, Ann. Rev. Neurosci., 1 (1978) 327-361. Zuckerman, J. E., Herschman, H. R. and Levine, L., Appearance of a brain specific antigen (the S-100 protein) during human fetal development, J. Neurochem., 17 (1970) 247-251.
337
Differentiation of Purified Astrocytes in a Chemically Defined Medium RICHARD S. MORRISONL4 and JEAN DE VELLIS 1-5
Departments of tAnatomy and 2Psychiatry, School of Medicine, 3MentalRetardation Research Center, 4Laboratoryof Biochemical and Environmental Sciences, and 5Brain ResearchInstitute, Universityof California, Los Angeles, CA 90024(U.S.A.) (Accepted February 22nd, 1983)
Key words: astrocytes - - growth factors - - differentiation - - glial fibrillary acidic protein - - CNS - - cell culture
Homogeneous cultures of astrocytes and oligodendrocytes provide an excellent model system for studying the regulation of glial structure and function. Recently, a chemically defined (CD) medium was developed for purified cultures of astrocytes, thus eliminating the requirement for serum and providing a controlled system for the study of astroglial properties. Due to the widespread use of astrocyte cultures and the potential benefits to be gained from using a defined medium, astrocyte cultures raised in CD medium were analyzed for purity as well as morphological and biochemical properties. Purity was assessed using immunocytochemical staining for glial fibrillary acidic protein (GFAP) and fibronectin. Astrocytes raised in CD medium are 95% pure using the expression of GFAP as a criterion. Fewer than 1% of the cells in CD medium stained positive for fibronectin eliminating the possibilitythat CD medium is selective for meningeal or endothelial cells. Astrocytes raised in CD medium exhibit a striking degree of morphological differentiation as seen in scanning electron micrographs. They also exhibit a high degree of biochemical differentiation illustrated by increases in the specific activity of S-100 protein and the induction of glutamine synthetase by glucocorticoids. A defined medium that supports the proliferation of rat astrocytes and enhances numerous morphological and biochemical properties should greatly facilitate the study of factors controlling glial proliferation and differentation.
INTRODUCTION To elucidate the mechanisms governing the develo p m e n t of the nervous system it is necessary to identify factors which regulate the proliferation and differentiation of neural cells. T h e search for substances which are capable of influencing neural d e v e l o p m e n t has met with some success in investigations of both the PNS 6,13,18,24--26,32 and CNS 9,23,29. These advances in n e u r o b i o l o g y have been greatly aided, in part, by the d e v e l o p m e n t of in vitro cell culture systems. This is especially true for the PNS from which it is possible to p r o d u c e highly purified cultures of neurons and supporting cells, Progress in the analysis of central neurogenesis has been h i n d e r e d , however, due to difficulties in p r e p a r i n g purified cultures of the various neural cell types. A l t h o u g h a m e t h o d is still not available for p r o d u c i n g h o m o g e n o u s cultures of CNS neurons, it is now possible to g e n e r a t e highly purified cultures of astrocytes and o l i g o d e n d r o c y t e s 16. Previous studies on astroglial proliferation and differentiation have utilized fetal calf serum as a component in the culture mediumSA2,16,20,28.30. The unde0165-3806/83/$03.00 © 1983 Elsevier Science Publishers B.V.
fined nature of serum and the variability of many of its c o m p o n e n t s from one batch to a n o t h e r was a complicating factor in these studies, reducing the control one has over environmental factors in cell culture. In a recent communication 17 we r e p o r t e d that the requirement for serum could be circumvented by growing purified astrocytes in a chemically defined (CD) medium. W e now r e p o r t that astrocytes raised in C D m e d i u m exhibit e n h a n c e d morphological and biochemical differentiation. MATERIALS AND METHODS
Cell culture Purified cultures of rat astrocytes were p r e p a r e d from p r i m a r y cerebral cultures by the m e t h o d of McCarthy and de Vellis 16. Meningeal cell cultures were p r e p a r e d by removing the meninges from the cortices of 0 to 1-day-old rat pups, and dissociating the tissue with trypsin for 30 min. Trypsinization was stopped by adding culture m e d i a s u p p l e m e n t e d with 10% FCS. The cell suspension was centrifuged for 5 min at 800 rpm in an International Clinical Centrifuge. The
338 supernatant was discarded and the meningeal cells were resuspended in fresh serum-supplemented medium for plating. Both astroglial and meningeal primary cell cultures were initially plated in Falcon plastic tissue culture flasks (75 cm2), and maintained in Hams F-12 medium/Dulbecco-Vogt modification of Eagles medium (1:1 v/v), with 1.2 g/l of NaHCO 3, 15 mM hepes buffer (serum free (SF) medium) and 10% (v/v) FCS. Purified cultures of astrocytes and meninges were trypsinized with a 1% Enzar-T (40x stock concentrate) trypsin concentrate and 0.1 mM EDTA in Hank's balanced salt solution (Ca 2+- and Mg2+-free) and then diluted 1:4 in medium containing 10% FCS. The cells were counted in a Royco 927T cell tissue counter and plated out at the appropriate densities. Cells were plated directly on plastic unless stated otherwise. All cell cultures were maintained in a humidified atmosphere of 5% CO2-95% air at 37 °C.
CD medium This consisted of the SF medium plus 50 mM hydrocortisone, 100 pM putrescine, prostaglandin F2a (PGF2,, 500 ng/ml), insulin (50 ~g/ml), and Fibroblast Growth Factor (FGF, 100 ng/ml).
dilution), washed extensively, and then stained with fluorescein isothiocyanate (FITC)-conjugated swine anti-rabbit IgG (1:50 dilution; 30 min). After additional washing the coverslips were rinsed in distilled water and mounted on glass slides with glycerol/phosphate-buffered saline, 1:9 (v/v). Cells were observed through a Zeiss fluorescence microscope equipped with FITC filters and mercury-vapor epi-illumination.
Scanning electron microscopy (SEM) Cultures were rinsed twice with physiological saline (0.85%), and then fixed with 4% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4, 0.05% calcium chloride added) for 2 h at room temperature. Following the rinse with buffer, cultures were postfixed with 1% osmium tetroxide in the same cacodylate buffer for 2 h. Cultures were then dehydrated in graded ethyl alcohols. After dehydration, the specimens were critically point dried from CO2. They were coated for 2 min using 15 s intervals with gold palladium in a Hummer I coater. Cells were observed with an ETEC Autoscan microscope.
Enzyme assays Glial fibrillary acidic protein (GFAP) and fibronectin immunofluorescence Astroglial and meningeal cell cultures were grown on 15 mm glass coverslips contained in microwell plates (Falcon). Astrocytes were seeded at 1 x 104 per well (2.1 cm2) in serum-supplemented medium for 18 h, rinsed 3× with SF medium and then switched to CD medium. Meningeal cells were seeded at 2 × 104 per well (2.1 cm 2) in serum-supplemented medium for 48 h, rinsed 3 times with SF medium and then switched to CD medium. Control cultures were maintained in serum-supplemented medium for the duration of the experiment. After 7 or 14 days in CD medium (7 days for meningeal cell cultures) cell cultures were washed with phosphate-buffered saline (pH 7.4) and fixed at room temperature for 15 rain in 2% (w/v) paraformaldehyde (0.1 M sodium cacodylate buffer/0.05% calcium chloride, pH 7.4). Cultures used for the study of GFAP were fixed for an additional 5 min with acetone. The cells were incubated for 30 rain with anti-GFAP rabbit antiserum (1:200 dilution), or anti-fibronectin rabbit antiserum (1:160
The methods used to assay glycerol phosphate dehydrogenase (GPDH) and glutamine synthetase (GS) have been previously described2,22. One unit of enzyme activity is defined as the transformation of one nmole of substrate per min at 30 °C for GPDH and 37 °C for GS. Specific activity is expressed as units of enzyme activity per mg of total cell protein. All assays were performed in triplicate.
S-IO0 protein Values of S-100 protein were determined as described by Zuckerman et al. 33 by the micro-complement fixation of LevineU.
Protein assay Protein content was determined by the method of Lowry et al.~4 using crystallized bovine serum albumin as a standard.
Materials Materials were obtained from the following sources: crystalline bovine insulin, PGF2a, and putrescine
339 dihydrochloride from Sigma; FGF from Collaborative Research; hydrocortisone acetate from Calbiochem-Behring; Ham's F-12, DMEM, and Hank's Balanced Salt Solution (Mg 2÷- and Ca2+-free) from Gibco; fetal calf serum from Irvine Scientific; trypsin from Reheis; fluorescein conjugated swine anti-rabbit IgG from Bio-Rad Laboratories; guinea pig complement and sheep red blood cells from M. A. Bioproducts; hemolysin from BBL; division of Becton, Dickinson; and L-[U-14C]glutamic acid (285 mCi/~mol) from Amersham. The following gifts are gratefully acknowledged: anti-GFAP rabbit antiserdm supplied by Dr. Larry F. Eng (Stanford University and V.A. Hospital, Palo Alto), and S-100, anti-S-100 rabbit antiserum and antifibronectin rabbit antiserum supplied by Dr. Harvey R. Herschman (UCLA).
RESULTS Cultures of purified astrocytes raised in CD medium were evaluated with respect to their purity and state of differentiation. The differences between astrocyte cultures raised in the presence of fetal calf serum or in defined medium have not been previously reported. To address this question, we grew purified astrocytes in parallel cultures under these two conditions. Astrocytes grown in serum-supplemented medium (10% FCS) for one or two weeks appear flat and morphologically undifferentiated (Fig. 1A). These cells possess GFAP as demonstrated by indirect immunofluorescence (Fig. 1B). An analysis of randomly chosen fields on several tissue culture dishes showed that 90% of the cells specifically stained for GFAP
Fig. 1. GFAP immunofluorescence of astrocyte cultures grown in serum-supplemented medium (10%). The same field from a representative culture was visualized by phase-contrast microscopy(A) and indirect immunofluorescencefor GFAP (B). Substitution of
GFAP-antisera by normal rabbit serum IgG displayed no visible fluorescence. All other contr*ols,including blocking the antiserum with purified GFAP and reagent controls, were also negative (x 520).
340 TABLE I Temporal expression of'cell markers
Imm~mocytochemistrywas performed as described in Materials and Methods. The data represent the percentage of cells expressing the specific marker in relation to the total cell population. The value of S-100 protein at 2 weeks in culture was significantly greater in CD medium than in serum-supplement medium (P ~<0.001). 1 Week
GFAP Fibronectin S-100 ~g/mg protein)
2 Weeks
Serum
CDM
Serum
CDM
>90.00% < 5.00% 0.40+0.06
>95.00% < 1.00% 0.25+0.06
>90.00% < 5.00% 1.20+0.09
>95.00% < 1.00% 1.42+0.07
(Table I). Astrocytes expressing G F A P exhibited fluorescence in the soma and processes. Nuclei were never observed to fluoresce. The cells raised in CD medium were 95% astrocytic as judged by their expression of G F A P (Table I). The cellular distribution of G F A P was the same as that observed for cells in serum-supplemented medi-
um, but the immunofluorescence clearly accentuated the enhanced formation of processes observed in cells grown in CD medium (Fig. 2B). In addition to the greater degree of cellular processes, G F A P immunofluorescence was much more intense in cells grown in CD medium. Astrocyte cultures raised in CD medium were also
Fig. 2. GFAP immunofluorescence of astrocyte cultures grown in CD medium. The same field from a representative culture was visualized by phase-contrast microscopy (A) and indirect immunofluorescence for GFAP (B). For controls, see Fig. 1 (x 520).
341
Fig. 3. Fibronectin immunofluorescence of meningeal cell cultures. Meningeal cells were plated in serum-supplemented medium (10%, 48 h) and converted to CD medium for 7 days. The same field from a representative area was visualized by phase-contrast microscopy (A) and indirect immunofluorescence for fibronectin (B) (x 520).
examined for the appearance of fibronectin. Fibronectin is considered a reliable marker of meningeal and endothelial cell populations27, 30. A random sampling of different culture fields did not show appreciable staining for fibronectin utilizing indirect immunofluorescence (data not shown). The percentage of fibronectin-positive cells was less than 1% (Table I). In contrast, cultures raised in serum-supplemented medium exhibited occasional loci of fibronectin-positive cells (data not shown). The percentage of cells expressing fibronectin in these cultures never amounted to more than 5% of the total cell population. Although the difference between the number of fibronectin-positive cells found in serum-supplemented and CD medium is not great under normal conditions, it can be accentuated. The meninges are normally removed when primary cultures of cerebrum are established. If this step is not carefully performed the percentage of fibronectin-positive cells is drastically increased. In cultures prepared in this manner they can represent as much as 36% and 20% of the total cell population in serum-supplemented and defined medium cultures, respectively. The near complete absence of fibronectin in cul-
tures raised in CD medium cannot be attributed solely to the removal of serum. Purified meningeal cultures were established as controls. Meningeal cells plated and raised in serum-supplemented medium display a pericellular matrix of fibronectin (data not shown). If the cells are plated in serum-supplemented medium for 2 days, and then shifted to the defined medium for 7 days the fibronectin matrix is still present (Fig. 3). The shift to defined medium did not visibly alter the structure of the fibronectin matrix. Astrocytes maintained in serum-supplemented medium are flat and possess few processes (Fig. 4A,C). In addition, their surfaces are smooth and devoid of any distinguishable features. These characteristics also apply to astrocyte cultures maintained in SF medium after the initial 18 h serum-preincubation (data not shown). Astrocytes maintained in CD medium, however, exhibited dramatic changes in cell morphology. The majority of cells exhibit small ovoid somas and long branching processes (Fig. 4B,D). A small population of cells, however, possessing the flat, polugonal shaped morphology observed in serum-supplemented cultures remains in cultures grown in CD medium.
342
C
D
Fig. 4. Scanning electron micrographs of astrocyte cultures raised in serum-supplemented medium (A) and CD medium (B) (x 1680). Astrocytes were plated directly into serum-supplemented or CD medium and maintained for 4 days before processing. Sections of Araldite-embedded cultures (cut perpendicular to plating surface and stained with Toluidine blue) are shown for cells grown in serumsupplemented (C) and CD medium (D). Cells in CD medium display numerous filopodia (B, arrow). Bar equals 5 #m.
The somas of process-bearing cells possessed many convoluted folds and the processes exhibited evenly spaced arrays of filopodia (Fig. 4B). The tiMpodia were not present where processes appeared to be forming in the transition from a flattened morphology to the process-bearing morphology. These flattened segments clearly resemble the morphology of cells grown in serum-supplemented medium. The most notable difference between processbearing cells was the variable number of processes found issuing forth from the cell body. Astrocytes grown in CD medium exhibit not only a striking degree of morphological differentiation but
also a high degree of biochemical differentiation. Aside from expressing G F A P several other properties appear to be enhanced in these cells (Table I). S-100 protein, a defined glial marker7 increased 3- and 6-fold from week 1 to week 2 in serum-supplemented and CD medium, respectively. S-100 protein is increased from 1.20 ~g/mg protein in serum-supplemented cultures to 1.42/~g/mg protein in CD medium at week two. S-100 protein was not detected, however, in control cultures of rat oligodendrocytes, bovine endothelial cells and rat skin fibroblasts (data not shown). In the CNS, the enzyme glutamine synthetase (GS) is also restricted in location to glial
343
~_ t4-A' +Serum
most commonly employed. Unfortunately, the inclusion of serum has several disadvantages. Serum contains many undefined components, some of which may actually inhibit cell growth or interact with substances being studied. Furthermore, defined serum components commonly exhibit variation from batch to batch thereby influencing experimental reproduc-
B CDM
8
i
GS
fil
GPDH
n n= GS
GPDH
Fig. 5. Enzyme assays were performed as described in Materials and Methods. To measure the induction of GS, astrocytes were plated at 105 cells/35 mm dish in serum-supplemented medium. Eighteen hours later, half of the dishes were converted to CD medium minus hydrocortisone. These conditions were maintained for 3 days at which time cultures in both serum-supplemented and CD medium were induced for 48 h, with 1 pm hydrocortisone. All the media were replenished after the first 24 h incubation. Hydrocortisone induction of glutamine synthetase in CD medium was significantlygreater than the induction observed in serum-supplemented medium (P < 0.01). There was no significant induction of GPDH by hydrocortisone under either conditions.
cells ~5 and has been shown to be regulated by glucocorticoids 28. In CD medium, hydrocortisone (1/~M) induced GS activity approximately 4-fold over control cultures, compared to only a 1.5-fold induction in serum-supplemented medium (Fig. 5). Basal levels of GS activity were substantially increased in response to the presence of 50 nM hydrocortisone which was included as part of the growth medium (data not shown). Glycerol phosphate dehydrogenase (GPDH) is another glial enzyme regulated by glucocorticoids 2. The induction of G P D H in neural tissue appears to be restricted to oligodendrocytes ~0. In accordance with this observation we observed that although basal levels of G P D H are detectable, G P D H is not inducible in astrocyte cultures maintained in either CD or serum-supplemented medium (Fig. 5). DISCUSSION Dissociated cultures of neural tissue provide an excellent model system for studying neural structure and function. Cells in culture, however, have traditionally required an organ extract of one type or another as a supplement for plating and proliferation. Serum, usually from fetal animals, is the supplement
ibility. In order to more effectively study neural cells in culture it would be useful to define their nutritional and hormonal needs. A significant start can be made by removing serum from the growth medium. We approached this goal by developing a CD medium for purified cultures of astrocytes. The kinetics of astrocyte proliferation in the CD medium have been previously described 17. Astrocyte cultures raised in CD medium possess more GFAP-positive cells than sister cultures raised in serum-supplemented medium. The difference is not usually significant, but the trend has been observed in all of our experiments. More striking, however, is the difference in the intensity of staining for GFAP. GFAP immunofluorescence is much more intense in astrocyte cultures grown in CD medium. Although immunofluorescence is not a quantitative method it suggested that astrocytes raised in CD medium might possess greater quantities of GFAP. Recent data obtained by R I A demonstrates that astrocytes raised in CD medium contain 2-fold more GFAP (spec. act.) than control cultures raised in serum-supplemented medium (Morrison, Eng and de Vellis, unpublished observations). Astrocyte cultures rarely exhibited contamination by meningeal or endothelial cells as judged by the paucity of fibronectin immunofluorescence. Nevertheless, when fibronectin-positive cells were present they were much more common in cultures grown in serum-supplemented medium. This increase may reflect the incorporation of serum-fibronectin into a pericellular matrix already surrounding astrocytes. Several workers have demonstrated that exogenous fibronectin may be incorporated into the extracellular matrices of cells in culturea,L A more likely explanation, however, is that meningeal and endothelial cells require one or more serum components for survival and proliferation which are lacking in the defined medium. Several other reports on the use of defined medium mention a reduction in contamination by fibroblastic elements1.31. The reduction of fibro-
344 nectin immunofluorescence in our cultures is not due to the removal of serum. Meningeal cells maintained for 7 days in defined medium still display a fibronectin matrix that is not visibly altered from the matrix observed in serum-supplemented cultures. The most salient difference between astrocytes raised in serum-supplemented or C D medium is the dramatic change in morphology. In serum-supplemented medium astrocytes presented a fiat, undifferentiated, epithelioid morphology. In contrast, astrocytes raised in CD medium gradually extended long processes with evenly spaced arrays of filopodia, a morphology reminiscent of the mature astrocyte in vivo 19. A detailed study of the effects of the individual supplements on morphology has not been done. The interplay between all or several of the factors may be necessary for this morphological maturation. It is clearly not the result of simply removing serum, which has no long-term effect on the morphology of these cells. It is interesting to note that soluble extracts from brain can also stimulate the morphological maturation of glial cellsl2, 20. These substances may be growth factors similar to the components found in the defined medium. In support of this notion it has been reported that these morphological maturation factors also possess mitogenic activityS, 20. Astrocytes raised in CI~ medium express biochemical properties that are either maintained or enhanced with respect to cultures raised in the presence of fetal calf serum. S-100 protein for example, is significantly elevated (approximately 18%) in astrocyte REFERENCES 1 Barnes, D. and Sato, G., Methods for growth of cultured cells in serum-free medium, Analyt. Biochem., 102 (1980) 255-270. 2 Breen, G. A. M. and de Vellis, J., Regulation of glycerol phosphate dehydrogenase by hydrocortisone in dissociated rat cerebral cell cultures, Develop. Biol., 41 (1974) 255-266. 3 Gysin, R., Moore, B. W., Proffitt, R. T., Devel, T. F., Caldwell, K. and Glaser, L., Regulation of the synthesis of S-100 protein in rat glial cells, J. biol. Chem., 255 (1980) 1515-1519. 4 Hayman, E. G. and Ruoslahti, E., Distribution of fetal bovine serum fibronectin and endogenous rat cell fibronectin in extracellular matrix, J. Cell Biol., 83 (1979) 255-259. 5 Hayman, E. G., Engvall, E. and Ruoslahti, E., Concomitant loss of cell surface fibronectin and laminin from transformed rat kidney cells, J. Cell Biol., 88 (1981) 352-357. 6 Hendry, I. A., Cell division in the developing sympathetic nervous system, J. Neurocytol., 6 (1977) 299-309.
cultures raised in CD medium. The regulation of S-100 protein is not understood, but the data that are available are in part contradictory3,2k While it is agreed that S-100 protein is regulated by homotypic cellular contacts in C6 glioma cells, the effect(s) of peripheral agents such as hormones and growth factors is controversial. The development of a CD medium that promotes attachment, proliferation and differentiation of astrocytes marks a significant advance in our ability to study astrocytes in vitro. The defined conditions should facilitate the study of the molecular mechanisms controlling astroglial proliferation and differentiation, which will hopefully allow us to rectify controversies like the one surrounding the regulation of S-100 protein. ACKNOWLEDGEMENTS This work was supported by the Department of Energy Contract DE-AM03-76-SF00012, Cellular and Molecular Biology U.S. Public Health Service Training Grant GM07185, and U.S. Public Health Service Grants HD-05615 and AG-01754. We wish to thank Dr. Alaric T. A r e n a n d e r for critically reading the manuscript and Dr. Peter Shintaku for his expert assistance with the scanning electron microscopy and illustrations. We also thank Joyce Adler and Mary A n n Cook for their help in preparing the manuscript.
7 Hyden, H. and McEwen, B. S., A glial protein specific for the nervous system, Proc. nat. Acad. Sci. U.S.A., 55 (1966) 354-358. 8 Kato, T., Fukui, Y., Turriff, D. E., Nakagawa, S., Lim, R.. Arnason, B. G. W. and Tanaka, Glial maturation factor in bovine brain: partial purification and physicochemical characterization, Brain Res., 212 (1981) 393--402. 9 Landmesser, L. and Pilar, G., Interactions between neurons and their targets during in vivo synaptogenesis, Fed. Proc., 37 (1978) 2016-2022. 10 Leveille, P. J., McGinnis, J. F., Maxwell, D. S. and de Vellis, J., Immunocytochemical localization of glycerol-3phosphate dehydrogenase in rat oligodendrocytes, Brain Res., 196 (1980) 287-305. 11 Levine, L., Microcomplement fixation. In D. M. Weir (Ed.), Handbook of Experimental lmmunochemistry, Blackwell Scientific Publications, Oxford, 1967, pp. 707-719. 12 Lim, R., Glia maturation factor, Current Topics in Develop. Biol., 16 (1980) 305-322. 13 Lofberg, J., Ahlfors, K. and Fallstrom, C., Neural crest cell
345
14
15
16
17
18
19
20
21
22
23
24
migration in relation to extracellular matrix organization in the embryonic Axolotl trunk, Develop. Biol., 75 (1980) 148-167. Lowry, O. H., Rosebrough, W. F., Farr, A. C. and Randall, R. V., Protein measurement with the folin phenol reagent, J. biol. Chem., 193 (1951)265-275. Martinez-Hernandez, A., Bell, K. P. and Norenberg, M. D., Glutamine synthetase: glial localization in brain, Science, 195 (1356-1358). McCarthy, K. D. and de Vellis, J., Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue, J. Cell Biol., 85 (1980) 890-902. Morrison, R. S. and de Vellis, J., Growth of purified astrocytes in a chemically defined medium, Proc. nat. Acad. Sci. U.S.A., 78 (1981) 7205-7209. Patterson, P. H., Environmental determination of autonomic neurotransmitter functions, Ann. Rev. Neurosci., 1 (1978) 1-17. Peters, A., Palay, S. L.and Webster, H. def., The Fine Structure of the Nervous System: The Neurons and Supporting Cells, W. B. Saunders, Philadelphia, 1976. Pettman, B., Sensenbrenner, M. and Labourdette, G., Isolation of a glial maturation factor from beef brain, FEBS Lett., 118 (1980) 195-199. Pfeiffer, S. E., Herschman, H. R., Lightbody, J. and Sato, G., Synthesis by a clonal line of rat glial cells of a protein unique to the nervous system, J. cell. Physiol., 75 (1970) 329-339. Pishak, M. R. and Phillips, A. T., A modified radioisotopic assay for measuring glutamine synthetase activity in tissue extracts, Analyt. Biochem., 94 (1979) 82-88. Roger, L. J. and Fellows, R. E., Evidence for thyroxinegrowth hormone interaction during brain development, Nature (Lond.), 282 (1979) 414--415. Salzer, J. L. and Bunge, R. P., Studies of Schwann prolifer-
25
26
27
28
29
30
31
32
33
ation during development, Wallerian degeneration, and direct injury, J. Cell Biol., 84 (1980) 739-752. Salzer, J. L., Williams, A. K., Glaser, L. and Bunge, R. P., Studies of Schwann cell proliferation. II, Characterization of the stimulation and specificity of the response to a neurite membrane fraction, J. Cell Biol., 84 (1980) 735-766. Salzer, J. L., Bunge, R. P. and Glaser, L., Studies of Schwann cell proliferation. III. Evidence for the surface localization of the neurite mitogen, J. Cell Biol., 84 (1980) 767-768. Schachner, M. G., Schoonmaker, G. and Hynes, R. O., Cellular and subcellular localization of LETS protein in the nervous system, Brain Res., 158 (1978) 149-158. Schousboe, A. G., Svenneby, G. and Hertz, L., Uptake and metabolism of glutamate in astrocytes cultures from dissociated mouse brain hemispheres, J. Neurochem., 29 (1977) 999-1005. Smith, R. M., Patel, A. J., Kingsbury, A. E., Hunt, A. and Balazs, R., Effects of thyroid state on brain development: beta-adrenergic receptors and 5'-nucleotidase activity, Brain Res., 198 (1980) 375-387. Steig, P. E., Kimelberg, H. K., Mazurkiewicz, J. E. and Banker, G. A., Distribution of glial fibrillary acidic protein and fibronetin in primary astroglial cultures from rat brain, Brain Res., 199 (1980) 493-500. Taub, M. and Sato, G., Growth of functional primary cultures of kidney epithelial cells in defined medium, J. Cell Physiol., 105 (1980) 369-378. Varon, S. S. and Bunge, R. P., Trophic mechanisms in the peripheral nervous system, Ann. Rev. Neurosci., 1 (1978) 327-361. Zuckerman, J. E., Herschman, H. R. and Levine, L., Appearance of a brain specific antigen (the S-100 protein) during human fetal development, J. Neurochem., 17 (1970) 247-251.