Proliferation of glial-derived cells in defined media

Factors influencing astrocyte growth and development in defined media

A previously described serum-free, defined medium (G2 medium) containing transferrin, selenium, hydrocortisone, biotin, fibroblast growth factor (FGF) and fibronectin developed for the growth of human and rat derived glioma cells was investigated for its ability to support proliferation of astrocytes in primary cultures of neonatal rat cerebrum. These cells were able to grow in G2 medium. Enhanced proliferation and repeated subcultivation were obtained after adding insulin and/or epidermal growth factor (EGF) to the G2 medium at concentrations of 5 μg/ml and 10 ng/ml, respectively. In these modified media (called G4 and G5 medium) astrocytes showed a higher degree of morphological differentiation as compared to serum supplemented medium. Cell type specificity was determined by immunocytochemical staining of glial fibrillary acidic (GFA) protein, which could already be demonstrated 5 days after plating cells. G4 and G5 represent the first serum-free defined media in which astrocytes proliferate and differentiate without preceding or intermediate contact to serum supplemented medium. Modification of the culture substratum by adding hyaluronic acid and chondroitin sulfate A to G4 medium (G2 medium + insulin) enhanced proliferation of astroglial cells by a factor of about 1.5. In the presence of epidermal growth factor no response to the altered culture dish surface was observed and the addition of fibronectin, otherwise a stringent plating requirement, was no longer necessary.

Chemically defined medium for rat astroglial cells in primary culture

We have developed a serum-free defined medium that supports the growth in primary culture of rat astroglial cells. Cells dissociated from cerebral hemispheres of newborn rats were maintained for 4 days in a basal medium (Waymouth's medium) containing 10% fetal calf serum, which was substituted by a serum-free medium. The basal medium was supplemented with insulin (5 μg/ml) and fatty acid free bovine serum albumin (0.5 mg/ml). Under these conditions the cells proliferate as estimated by cell counts and DNA content; however, growth was less than in Waymouth's medium supplemented with 10% fetal calf serum. In contrast, a very similar morphology was observed between cultures grown in the serum-free or serum-containing media. The serum-free medium allows some maturation of the astroglial cells as shown by the presence of glial fibrillary acidic (GFA) protein, S-100 protein and glutamine synthetase (GS) activity. The astroglial cells can survive and grow in this chemically defined medium for up to 5 weeks. The ability to culture astroglial cells in such a minimal defined medium should facilitate investigations concerning the effects of growth factors on their proliferation and maturation.

Motoneuron programmed cell death in response to proBDNF

Motoneurons (MN) as well as most neuronal populations undergo a temporally and spatially specific period of programmed cell death (PCD). Several factors have been considered to regulate the survival of MNs during this period, including availability of muscle-derived trophic support and activity. The possibility that target-derived factors may also negatively regulate MN survival has been considered, but not pursued. Neurotrophin precursors, through their interaction with p75(NTR) and sortilin receptors have been shown to induce cell death during development and following injury in the CNS. In this study, we find that muscle cells produce and secrete proBDNF. ProBDNF through its interaction with p75(NTR) and sortilin, promotes a caspase-dependent death of MNs in culture. We also provide data to suggest that proBDNF regulates MN PCD during development in vivo.

Hydrocortisone stimulates neurite outgrowth from mouse retinal explants by modulating macroglial activity

PURPOSE

There is mounting evidence that retinal ganglion cells (RGCs) require a complex milieu of trophic factors to enhance cell survival and axon regeneration after optic nerve injury. The authors' goal was to examine the contribution of components of a combination of hormones, growth factors, steroids, and small molecules to creating a regenerative environment and to determine if any of these components modulated macroglial behavior to aid in regeneration.

METHODS

Postnatal day 7 mouse retinal explants embedded in collagen were used as an in vitro model of neurite regeneration. Explants were treated with the culture supplements fetal bovine serum, N2, and G5 and a mixture of G5 and N2 components, designated enhanced N2 (EN2). Explants were evaluated for neurite outgrowth over 7 days in culture. The effects of each treatment were also evaluated on cultured RGCs purified by Thy1 immunopanning. Immunohistochemistry and qPCR analysis were used to evaluate differences in gene expression in the explants due to different treatments.

RESULTS

EN2 stimulated significant neurite outgrowth from explants but not from purified RGCs. Elimination of hydrocortisone (HC) from EN2 reduced the mean neurites per explant by 37%. EN2-treated explants demonstrated increased expression of Gfap, Glul, Glt1, Cntf, Pedf, and VegfA compared with explants treated with EN2 without HC. Subsequent experiments showed that increased expression of Cntf and Glul was critical to the trophic effect of HC.

CONCLUSIONS

These data suggest that the HC in EN2 indirectly contributed to neurite outgrowth by activating macroglia to produce neurotrophic and neuroprotective molecules.

The effects of orexins on monoaminerg-induced changes in vasopressin level in rat neurohypophyseal cell cultures

The effects of orexin-monoaminergic compound interactions on vasopressin release were studied in 14-day neurohypophyseal cell cultures from adult rats, prepared by an enzymatic dissociation technique. The vasopressin contents of the supernatants were determined by radioimmunoassay. Following administration of either orexin-A or orexin-B in increasing doses, significant changes were not observed in the vasopressin levels of the supernatant media. The vasopressin level substantially increased after epinephrine, norepinephrine, serotonin, histamine, dopamine or K(+) treatment. Preincubation with either orexin-A or orexin-B reduced the epinephrine-, histamine- or serotonin-induced increases in vasopressin level, but the vasopressin concentrations of the supernatant media remained above the control level. There was no significant difference in decreasing effect between orexin-A and orexin-B. Neither orexin-A nor orexin-B induced changes in vasopressin release following monoaminergic compound treatment. The results indicate that the changes in vasopressin secretion induced by the monoaminergic system can be directly influenced by orexin system. It may be presumed that the orexins can play a physiological role in the regulation of the water metabolism by reducing the effect of increased vasopressin release caused by monoaminergic compounds. The interactions between the monoaminergic and orexin systems regarding vasopressin secretion occur at both the hypothalamic and the neurohypophyseal level.

A two-step strategy for neuronal differentiation in vitro of human dental follicle cells

Human dental follicle cells (DFCs) derived from wisdom teeth are precursor cells for cementoblasts. In this study, we recognized that naïve DFCs express constitutively the early neural cell marker beta-III-tubulin. Interestingly, DFCs formed beta-III-tubulin-positive neurosphere-like cell clusters (NLCCs) on low-attachment cell culture dishes in serum-replacement medium (SRM). For a detailed examination of the neural differentiation potential, DFCs were cultivated in different compositions of SRM containing supplements such as N2, B27, G5 and the neural stem cell supplement. Moreover, these cell culture media were combined with different cell culture substrates such as gelatin, laminin, poly-L-ornithine or poly-L-lysine. After cultivation in SRM, DFCs differentiated into cells with small cell bodies and long cellular extrusions. The expression of nestin, beta-III-tubulin, neuron-specific enolase (NSE) and neurofilament was up-regulated in SRM supplemented with G5, a cell culture supplement for glial cells, and the neural stem cell supplement. DFCs formed NLCCs and demonstrated an increased gene expression of neural cell markers beta-III-tubulin, NSE, nestin and for small neuron markers such as neuropeptides galanin (GAL) and tachykinin (TAC1) after cultivation on poly-L-lysine. For a further neural differentiation NLCC-derived cells were sub-cultivated on laminin and poly-L-ornithine cell culture substrate. After 2 weeks of differentiation, DFCs exposed neural-like cell morphology with small neurite-like cell extrusions. These cells differentially express neurofilament and NSE, but only low levels of beta-III-tubulin and nestin. In conclusion, we demonstrated the differentiation of human DFCs into neuron-like cells after a two-step strategy for neuronal differentiation.

Effects of galanin-monoaminergic interactions on vasopressin secretion in rat neurohypophyseal cell cultures

The effects of dopamine (DA), serotonin (5-HT), histamine (HA), adrenaline (ADR), noradrenaline (NADR) and K(+) administration on vasopressin (VP) secretion were studied in 13-14-day cultures of rat neurohypophyseal (NH) cells, and it was examined whether galanin (GAL) can modify the VP release enhancement induced by these monoaminergic compounds. An enzymatic dissociation technique was used to make the rat NH cell cultures. The VP contents of the supernatants of 14-day cultures were determined by radioimmunoassay. Following the administration of 10(-6) M GAL, the VP secretion into the supernatant media decreased. DA, 5-HT, ADR or NADR treatment increased the VP level substantially, while the enhancing effect of HA was more moderate. GAL administration before DA, ADR and NADR treatment prevented the VP concentration increase induced by DA, ADR or NADR. Preincubation with GAL reduced the 5-HT- or HA-induced VP level increases; the VP concentrations of the supernatant media remained above the control level. The GAL blocking effect was prevented by previous treatment with the GAL receptor antagonist galantid (M15). GAL had no effect on the VP level increase induced by K(+), which causes a non-specific hormone secretion. The results indicate that the changes in VP secretion induced by the monoaminergic system can be directly influenced by the GAL-ergic system. The interactions between the monoaminergic and GAL-ergic systems regarding VP secretion occur at the level of the posterior pituitary.

Regulation of heat shock protein 70 release in astrocytes: role of signaling kinases

The ability to mount a successful stress response in the face of injury is critical to the long-term viability of individual cells and to the organism in general. The stress response, characterized in part by the upregulation of heat shock proteins, is compromised in several neurodegenerative disorders and in some neuronal populations, including motoneurons (MNs). Because astrocytes have a greater capacity than neurons to survive metabolic stress, and because they are intimately associated with the regulation of neuronal function, it is important to understand their stress response, so that we may to better appreciate the impact of stress on neuronal viability during injury or disease. We show that astrocytes subjected to hyperthermia upregulate Hsp/c70 in addition to intracellular signaling components including activated forms of extracellular-signal-regulated kinase (ERK1/2), Akt, and c-jun N-terminal kinase/stress activated protein kinase (JNK/SAPK). Furthermore, astrocytes release increasing amounts of Hsp/c70 into the extracellular environment following stress, an event that is abrogated when signaling through the ERK1/2 and phosphatidylinositol-3 kinase (PI3K) pathways is compromised and enhanced by inhibition of the JNK pathway. Last, we show that the Hsp/c70 is released from astrocytes in exosomes. Together, these data illustrate the diverse regulation of stress-induced Hsp/c70 release in exosomes, and the way in which the balance of activated signal transduction pathways affects this release. These data highlight how stressful insults can alter the microenvironment of an astrocyte, which may ultimately have implications for the survival of neighboring neurons.

Astrocyte and muscle-derived secreted factors differentially regulate motoneuron survival

During development, motoneurons (MNs) undergo a highly stereotyped, temporally and spatially defined period of programmed cell death (PCD), the result of which is the loss of 40-50% of the original neuronal population. Those MNs that survive are thought to reflect the successful acquisition of limiting amounts of trophic factors from the target. In contrast, maturation of MNs limits the need for target-derived trophic factors, because axotomy of these neurons in adulthood results in minimal neuronal loss. It is unclear whether MNs lose their need for trophic factors altogether or whether, instead, they come to rely on other cell types for nourishment. Astrocytes are known to supply trophic factors to a variety of neuronal populations and thus may nourish MNs in the absence of target-derived factors. We investigated the survival-promoting activities of muscle- and astrocyte-derived secreted factors and found that astrocyte-conditioned media (ACM) was able to save substantially more motoneurons in vitro than muscle-conditioned media (MCM). Our results indicate that both ACM and MCM are significant sources of MN trophic support in vitro and in ovo, but only ACM can rescue MNs after unilateral limb bud removal. Furthermore, we provide evidence suggesting that MCM facilitates the death of a subpopulation of MNs in a p75(NTR) - and caspase-dependent manner; however, maturation in ACM results in MN trophic independence and reduced vulnerability to this negative, pro-apoptotic influence from the target.

Potassium currents in primary cultured astrocytes from the rat corpus callosum

The corpus callosum (CC) is the main white matter tract in the brain and is involved in interhemispheric communication. Using the whole-cell voltage-clamp technique, a study was made of K(+)-currents in primary cultured astrocytes from the CC of newborn rats. These cells were positive to glial fibrillary acidic protein after culturing in Dulbecco's Modified Eagle Medium (> 95% of cells) or in serum-free neurobasal medium with G5 supplement (> 99% of cells). Astrocytes cultured in either medium displayed similar voltage-activated ion currents. In 81% of astrocytes, the current had a transient component and a sustained component, which were blocked by 4-aminopyridine and tetraethylammonium, respectively; and both had a reversal potential of -66 mV, indicating that they were carried by K(+) ions. Based on the Ba(2+)-sensitivity and activation kinetics of the K(+)-current, two groups of astrocytes were discerned. One group (55% of cells) displayed a strong Ba(2+) blockade of the K(+)-current whose activation kinetics, time course of decay, and the current-voltage relationship were modified by Ba(2+). This current was greatly blocked (52%) by Ba(2+) in a voltage-dependent way. Another group (45% of cells) presented weak Ba(2+)-blockade, which was only blocked 24% by Ba(2+). The activation kinetics and time course of decay of this current component were unaffected by Ba(2+). These results may help to understand better the roles of voltage-activated K(+)-currents in astrocytes from the rat CC in particular and glial cells in general.

Histamine-induced enhancement of vasopressin and oxytocin secretion in rat neurohypophyseal tissue cultures

The effects of histamine (HA) on vasopressin (VP) and oxytocin (OT) secretion were studied in 13-14-day cultures of isolated rat neurohypophyseal (NH) tissue. The VP and OT contents of the supernatant were determined by radioimmunoassay (RIA) after a 1 or 2-h incubation. Significantly increased levels of VP and OT production were detected in the tissue culture media following HA administration, depending on the HA dose. The elevation of NH hormone secretion could be partially blocked by previous administration of the HA antagonist mepyramine (MEP, an H1 receptor antagonist) or cimetidine (CIM, an H2 receptor antagonist). Thioperamide (TPE, an H3-H4 receptor antagonist) did not influence the VP or OT secretion increase induced by HA. The application of MEP, CIM or TPE after HA administration proved ineffective. The H1 and H2 receptors are mainly involved in the HA-induced increase of both VP and OT secretion in isolated NH tissue cultures. The results indicate that NH hormone release is influenced directly by the histaminergic system, and the histaminergic control of VP and OT secretion from the NH tissue in rats can occur at the level of the posterior pituitary.

Serotonin-induced enhancement of vasopressin and oxytocin secretion in rat neurohypophyseal tissue culture

The effects of serotonin (5-hydroxytryptamine; 5-HT) on vasopressin (VP) and oxytocin (OT) secretion were studied in 13-14-day cultures of isolated rat neurohypophyseal (NH) tissue. The VP and OT contents of the supernatant were determined by radioimmunoassay after a 1 or 2 h incubation. Significantly increased levels of VP and OT production were detected in the tissue culture media following 5-HT administration, depending on the 5-HT dose. The elevation of NH hormone secretion could be partially blocked by previous administration of the 5-HT antagonist ketanserin or metergoline. WAY-100635 did not influence the increased VP secretion induced by 5-HT, but the elevated OT production was prevented by WAY-100635 before 5-HT administration. The application of WAY-100635, ketanserin or metergoline, after 5-HT administration proved ineffective. The results indicate that NH hormone release is influenced directly by the serotonergic system. The serotonergic control of VP and OT secretion from the NH tissue in rats can occur at the level of the posterior pituitary.

Failure to express GAP-43 leads to disruption of a multipotent precursor and inhibits astrocyte differentiation

The nervous system-specific protein GAP-43 is significantly upregulated in neurons and glia that are differentiating. In P19 EC cells that do not express GAP-43, neurogenesis is inhibited; many immature neurons apoptose and the survivors do not mature morphologically. Here we show that the initial defect is in an early precursor with characteristics of a neural stem cell, which failed to respond normally to retinoic acid (RA). As a consequence, its progeny had altered cell fates: In addition to the neuronal defects previously reported, RC1-labeled radial glia failed to exit the cell cycle, accumulated, and failed to acquire GFAP immunoreactivity. However, leukemia inhibitory factor (LIF) could stimulate GFAP expression suggesting that astrocytes not derived from radial glia are less affected by absence of GAP-43. Differentiation of radial glia-derived astrocytes was also inhibited in glial cultures from GAP-43 (-/-) cerebellum, and in GAP-43 (-/-) telencephalon in vivo, differentiation of astrocytes derived from both radial and nonradial glia lineages were both affected: In the glial wedge, GFAP-labeled radial glia-derived astrocytes were reduced consistent with the interpretation that they may be unable to deflect GAP-43 (-/-) commissural axons toward the midline. At the midline, both radial and nonradial glia-derived astrocytes were also decreased although it fused normally. The results demonstrate that GAP-43 expressed in multipotent precursors is required for appropriate cell fate commitment, and that its absence affects astrocyte as well as neuronal differentiation.

Fibroblast growth factors redirect retinal axons in vitro and in vivo

Growth factors have been shown previously to participate in the process of axon target recognition. We showed that fibroblast growth factor receptor (FGFR) signaling is required for Xenopus laevis retinal ganglion cell (RGC) axons to recognize their major midbrain target, the optic tectum [neuron 17 (1996), 245]. Therefore, we have hypothesized that a change in expression of a fibroblast growth factor (FGF) at the entrance of the optic tectum, the border between the diencephalon and mesencephalon, may serve as a signal to RGC axons that they have reached their target. To determine whether RGC axons can sense changes in FGF levels, we asked whether they altered their behavior upon encountering an ectopic source of FGF. We found that in vivo RGC growth cones avoided FGF-misexpressing cells along their path, and that FGF-2 directly repelled RGC growth cones in an in vitro growth cone turning assay. These data support the idea that RGC axons can sense changes in FGF levels, and as such provide a mechanism by which FGFR signaling is involved in RGC axon target recognition.

Inhibitory effect of galanin on dopamine-induced enhanced vasopressin secretion in rat neurohypophyseal tissue cultures

The effect of galanin (GAL) on vasopressin (VP) secretion was studied in 13-14-day cultures of isolated rat neurohypophyseal (NH) tissue. The VP content of the supernatant was determined by radioimmunoassay (RIA) after a 1- or 2-h incubation. A significantly decreased content of VP was detected following the administration of 10(-6)-10(-9) M doses of GAL. Dopamine (DA) and the DA-active drugs apomorphine (APM) and Pro-Lys-Gly (PLG) (10(-6) M in each medium) increased the VP level of NH tissue cultures. This VP concentration elevation could be blocked by the administration of GAL together with DA, APM or PLG. The DA-blocking effect of GAL was prevented by previous treatment with the GAL receptor antagonist galantid (M15). The results indicate that VP release is directly influenced by the GAL-ergic system. The GAL-ergic control of VP secretion from NH tissue in rats can occur independently of the hypothalamus, at the level of the posterior pituitary.

Estrogen (E2) and glucocorticoid (Gc) effects on microglia and A beta clearance in vitro and in vivo

The accumulation of fibrillar aggregates of beta Amyloid (A beta) in Alzheimer's Disease (AD) brain is associated with chronic brain inflammation. Although activated microglia (mu glia) can potentially clear toxic amyloid, chronic activation may lead to excessive production of neurotoxins. Recent epidemiological and clinical data have raised questions about the use of anti-inflammatory steroids (glucocorticoids, Gcs) and estrogens for treatment or prevention of AD. Since very little is known about steroid effects on mu glial interactions with amyloid, we investigated the effects of the synthetic Gc dexamethasone (DXM) and 17-beta estradiol (E2) in vitro in a murine mu glial-like N9 cell line on toxin production and intracellular A beta accumulation. To determine whether the steroid alterations of A beta uptake in vitro had relevance in vivo, we examined the effects of these steroids on A beta accumulation and mu glial responses to A beta infused into rat brain. Our in vitro data demonstrate for the first time that Gc dose-dependently enhanced mu glial A beta accumulation and support previous work showing that E2 enhances A beta uptake. Despite both steroids enhancing uptake, degradation was impeded, particularly with Gcs. Distinct differences between the two steroids were observed in their effect on toxin production and cell viability. Gc dose-dependently increased toxicity and potentiated A beta induction of nitric oxide, while E2 promoted cell viability and inhibited A beta induction of nitric oxide. The steroid enhancement of mu glial uptake and impedence of degradation observed in vitro were consistent with observations from in vivo studies. In the brains of A beta-infused rats, the mu glial staining in entorhinal cortex layer 3, not associated with A beta deposits was increased in response to A beta infusion and this effect was blocked by feeding rats prednisolone. In contrast, E2 enhanced mu glial staining in A beta-infused rats. A beta-immunoreactive (ir) deposits were quantitatively smaller, appeared denser, and were associated with robust mu glial responses. Despite the fact that steroid produced a smaller more focal deposit, total extracted A beta in cortical homogenate was elevated. Together, the in vivo and in vitro data support a role for steroids in plaque compaction. Our data are also consistent with the hypothesis that although E2 is less potent than Gc in impeding A beta degradation, long term exposure to both steroids could reduce A beta clearance and clinical utility. These data showing Gc potentiation of A beta-induced mu glial toxins may help explain the lack of epidemiological correlation for AD. The failure of both steroids to accelerate A beta degradation may explain their lack of efficacy for treatment of AD.

Rapid morphological changes in astrocytes are accompanied by redistribution but not by quantitative changes of cytoskeletal proteins

Astrocytes have the potential to acquire very different morphologies, depending on their regional location in the CNS and on their functional interactions with other cell types. Morphological changes between a flat or a fibroblast-like and a stellate or process-bearing appearance, and vice versa, can occur rapidly, but very little is known as to whether morphological transformations are based on quantitative changes of cytoskeletal proteins in microfilaments, intermediate filaments, and/or microtubules. Using a cell culture of selective type 1 astrocytes, we compared the distribution and protein amounts of a number of cytoskeletal proteins both during primary process growth induced by specific media conditions and after secondary transformations induced by dBcAMP. Our data presented in this report support the idea that astrocytes can undergo dramatic changes in their morphology requiring subcellular redistribution of most cytoskeletal proteins but no quantitative modifications of the amount of the respective proteins. After pharmacological treatment with lysophosphatic acid and genistein we show that astrocytes can acquire intermediate morphologies reminiscent of both fibroblast and stellate-like cells. These experiments demonstrate that the recently described RhoA-mediated signaling cascade between the cell surface and cytoskeletal proteins is only one of several signaling pathways acting on the astrocytic cytoskeleton.

Effects of dopamine and dopamine-active compounds on oxytocin and vasopressin production in rat neurohypophyseal tissue cultures

The effects of dopamine (DA) or DA-active drugs on the synthesis of neurohypophyseal (NH) hormones were studied in 13-14 day cultures of isolated NH tissue from rats. The following DA-active compounds were used (10(-6) M in each medium): DA, apomorphine (APM), Pro-Lys-Gly (PLG), butaclamol (B), haloperidol (HP), chlorpromazine (CPZ) and sulpiride (SP). The oxytocin (OT) and vasopressin (VP) contents of the condensed media were determined by RIA after a 1 or 2 h incubation. Significantly increased contents of OT and VP were detected in the tissue culture media following DA, APM or PLG administration. This elevation of NH hormone production could be blocked by previous administration of B or the DA receptor antagonists HP, CPZ or SP. The application of B after DA agonists proved ineffective. The results indicate that NH hormone production can be directly influenced by the DA-ergic system. The DA-ergic control of NH hormone secretion in rats can occur independently of the hypothalamus, at the level of the posterior pituitary.

Heterogeneic modulation of malignant behavior in human glioma cells in defined and serum-containing media

Malignant features in three glioma cell lines were studied in four defined media of various complexity. The cell lines D37MG, D54MG, and GaMG were able to grow in monolayer culture in all media examined, and as multicellular tumor spheroids in the two most nutrient-rich media. In the defined media, none of the cell lines were able to migrate in a migration assay on poly-D-lysine-coated plastic surfaces. Flow cytometric analysis of the GaMG cell line demonstrated no medium-dependent selection of subclones of glioma cells in spheroids cultured for 30 d. Morphological diversity of spheroids varied according to the supplementation of the media. The capacity of glioma cells to invade cellular rat brain aggregates was intact in the media examined. However, glioma migration was severely inhibited by the lack of specific serum components. This study demonstrates that glioma growth and invasion was heterogeneously preserved in the defined media used. Depending on the assay to be used in the study of glioma cell behavior, the degree of medium supplementation has to be considered.

Adaptation of C6 glioma cells to serum-free conditions leads to the expression of a mixed astrocyte-oligodendrocyte phenotype and increased production of neurite-promoting activity

C6 cells were adapted to proliferate in defined culture medium to allow the study of long-term effects of serum-free growth conditions on their phenotypic antigen expression and production of neurite promoting factors (NPFs). Cultures were grown in either Ham's F-12 or supplemented Opti-MEM-I containing 15% heat-inactivated horse serum and 2.5% fetal calf serum (serum-containing) or in supplemented Opti-MEM-I alone (serum-free). Immunocytochemical and immunofluorescence techniques were used to determine the antigenic expression of A2B5, galactocerebroside (GalC), and glial fibrillary acidic protein (GFAP) in passage matched and sister cultures of serum and serum-free grown C6 cells. When C6 cells were grown under serum-containing conditions, two populations of cells were seen: young oligodendrocytes (A2B5+, GFAP-, GalC+), and mixed astrocyte-oligodendrocyte phenotype (A2B5+, GFAP+, GalC+). After adaptation of the C6 cells to serum-free conditions over 2-3 passages, only one population of cells was observed, the mixed astrocyte-oligodendrocyte phenotype. The serum-free conditions also resulted in greater staining of the C6 cells. Conditioned media from the two growth conditions were fractionated by ultrafiltration into two fractions: components > 50 kDa and components of 10-50 kDa. The amount of neurite promoting activity seen between the two culture conditions resulted in a 3-fold increase in NPF activity under serum-free conditions in the > 50 kDa fraction. The 10-50 kDa fraction only expressed NPF activity if obtained from the serum-grown C6 cells. This alteration in NPF activity appears to be the result of the phenotypical alteration of the C6 cells, and may suggest that the NPF activities from the two culture conditions may not be identical.

Growth factor effects on the proliferation of different retinal glial cells in vitro

Vascularized mammalian retinae contain two distinct neuroglial cells types, radially oriented Müller cells and astrocytes, which are located in the nerve fiber layer. These cell types derive from different precursor cells and proliferate during ontogenesis at distinct schedules. The aim of the present study was to disclose whether growth factors, which are known to interfere with the development of neuroglial cells in the central nervous system, like basic and acidic fibroblast growth factor (aFGF and bFGF), epidermal growth factor (EGF) and platelet-derived growth factor, have similar or distinct effects on the proliferative capacity of retinal astrocytes and Müller cells. These questions were tested by applying growth factors to cultured astrocytes and Müller cells from early postnatal rabbit retina. Proliferating cells were identified by double labeling experiments combining cell type specific markers with bromodeoxyuridine immunocytochemistry and [3H]thymidine incorporation experiments, respectively. In addition, we used the anatomical advantage of the rabbit retina. Its peripheral part is astroglial cell-free. Cultures prepared from this part of the retina (P-cultures) contain Müller cells, microglial cells and neurons, while cultures from the 'central part', the medullary rays (MR) region contain, in addition, astrocytes and oligodendrocytes. Our studies show that Müller cell proliferation is stimulated by EGF in a dose dependent manner, while astrocyte proliferation is stimulated by aFGF and bFGF. The proliferation of O4-positive glial precursor cells is stimulated by aFGF, bFGF and platelet-derived growth factor, but not by EGF. Microglial cells, which are a minor population in these cultures, do not respond to either of these factors.

Differential intracellular compartmentalization of phosphotyrosine phosphatases in a glial cell line: TC-PTP versus PTP-1B

Intracellular levels of protein-phosphotyrosine are regulated by the protein-tyrosine phosphatase (PTP) family. Cellular compartmentalization may play an important role in modulating the function of these enzymes. The recent demonstration that PTP-1B is localized to the endoplasmic reticulum (Frangioni et al: Cell 68:545, 1992) is consistent with this proposition. In this study we have examined the intracellular distribution of TC-PTP in a glial cell line (C6). Using indirect immunofluorescence we have shown that this enzyme is distributed differently from PTP-1B and is mainly concentrated in the perinuclear region of these cells.

Vitamin neurotoxicity

Vitamins contain reactive functional groups necessary to their established roles as coenzymes and reducing agents. Their reactive potential may produce injury if vitamin concentration, distribution, or metabolism is altered. However, identification of vitamin toxicity has been difficult. The only well-established human vitamin neurotoxic effects are those due to hypervitaminosis A (pseudotumor cerebri) and pyridoxine (sensory neuropathy). In each case, the neurological effects of vitamin deficiency and vitamin excess are similar. Closely related to the neurological symptoms of hypervitaminosis A are symptoms including headache, pseudotumor cerebri, and embryotoxic effects reported in patients given vitamin A analogs or retinoids. Most tissues contain retinoic acid (RA) and vitamin D receptors, members of a steroid receptor superfamily known to regulate development and gene expression. Vitamin D3 effects on central nervous system (CNS) gene expression are predictable, in addition to the indirect effects owing to its influence on calcium and phosphorus homeostasis. Folates and thiamine cause seizures and excitation when administered in high dosage directly into the brain or cerebrospinal fluid (CSF) of experimental animals but have rarely been reported to cause human neurotoxicity, although fatal reactions to i.v. thiamine are well known. Ascorbic acid influences CNS function after peripheral administration and influences brain cell differentiation and 2-deoxyglucose accumulation by cultured glial cells. Biotin influences gene expression in animals that are not vitamin-deficient and alters astrocyte glucose utilization. The multiple enzymes and binding proteins involved in regeneration of retinal vitamin A illustrate the complexity of vitamin processing in the body. Vitamin A toxicity is also a good general model of vitamin neurotoxicity, because it shows the importance of the ratio of vitamin and vitamin-binding proteins in producing vitamin toxicity and of CNS permeability barriers. Because vitamin A and analogs enter the CNS better than most vitamins, and because retinoids have many effects on enzyme activity and gene expression, Vitamin A neurotoxicity is more likely than that of most, perhaps all other vitamins. Megadose vitamin therapy may cause injury that is confused with disease symptoms. High vitamin intake is more hazardous to peripheral organs than to the nervous system, because CNS vitamin entry is restricted. Vitamin administration into the brain or CSF, recommended in certain disease states, is hazardous and best avoided. The lack of controlled trials prevents us from defining the lowest human neurotoxic dose of any vitamin. Large differences in individual susceptibility to vitamin neurotoxicity probably exist, and ordinary vitamin doses may harm occasional patients with genetic disorders.(ABSTRACT TRUNCATED AT 400 WORDS)

Response of malignant glioma cell lines to epidermal growth factor and platelet-derived growth factor in a serum-free medium

The use of a serum-free culture system for assessing the growth factor responsiveness of malignant glial cells is described. The mitogenic properties of epidermal growth factor (EGF) and platelet-derived growth factor (PDGF) were examined in three human malignant glioma cell lines (T98G, U87, and U138). Each of the three had high-affinity EGF receptors and all responded in a dose-dependent fashion to physiological concentrations of EGF. These cell lines also showed a pronounced mitogenic response to PDGF which equaled or exceeded that achieved with EGF. Simultaneous stimulation with both factors produced an additive response, which approximated that obtained in medium supplemented with 10% fetal calf serum. The authors conclude that functional EGF and PDGF receptors were present in the human malignant glial tumors studied. The response of the human glioma lines to these growth factors in many respects parallels the response seen in fetal astrocytes tested under similar conditions. In contrast, the behavior of two chemically induced rat gliomas (9L and C6) differed significantly from that seen in the human lines, suggesting that the rat lines may not be entirely acceptable as models for studying the growth characteristics of human malignant glial tumors.

Response of malignant glioma cell lines to activation and inhibition of protein kinase C-mediated pathways

To evaluate the role of protein kinase C-mediated pathways in the proliferation of malignant gliomas, this study examined the effect of a protein kinase C (PKC)-activating phorbol ester (12-O-tetradecanoyl-13-phorbol acetate or TPA) and a protein kinase C inhibitor (polymyxin B) on deoxyribonucleic acid (DNA) synthesis of malignant glioma cells in vitro. A serum-free chemically defined medium, MCDB 105, was employed for all studies. Two established human malignant glioma cell lines (T98G and U138), two rat glioma lines (9L and C6), and two low-passage human glioma lines (obtained from surgical specimens) were studied. With the exception of the C6 line, all tumors responded in a dose-dependent fashion to nanomolar concentrations of TPA with a median effective dose that varied from 0.5 ng/ml for the U138 glioma to 1 ng/ml for the T98G glioma. At optimal concentrations (5 to 10 ng/ml), TPA produced a two- to five-fold increase in the rate of DNA synthesis (p less than 0.05) as assessed by incorporation of 3H-thymidine. However, TPA had no additive effect on the mitogenic response produced by epidermal growth factor (EGF) or platelet-derived growth factor (PDGF). Inhibition of PKC using the antibiotic polymyxin B (20 micrograms/ml) abolished the TPA-induced mitogenic response in the five responsive lines tested. In two tumors (U138 and 9L), polymyxin B also eliminated EGF-, PDGF-, and serum-induced DNA synthesis as well as abolishing baseline DNA synthesis. These cells remained viable, however, as assessed by trypan blue exclusion; after removal of polymyxin B from the medium, they were able to resume DNA synthesis in response to TPA and serum. In the three other tumors (T98G and the two low-passage human glioma lines), growth factor-induced and serum-induced DNA synthesis were inhibited by approximately 25% to 85%. It is concluded that PKC-mediated pathways affect DNA synthesis in the human malignant glial tumors studied. The response of the glioma cells to TPA is similar to the responses seen in fetal astrocytes, but differs significantly from those reported for normal adult glial cultures. Because the response of the 9L glioma to TPA is similar to the responses seen in the human tumors, the 9L rat glioma model may prove useful for examining the role of PKC-mediated pathways in controlling glioma growth in vivo.

Suramin inhibits proliferation of rat glioma cells and alters N-CAM cell surface expression

Suramin, a drug used in the treatment of trypanosomiasis and onchocerciasis inhibits growth-factor-induced mitogenesis. We have investigated the effect of suramin on the growth rate and the morphology of C6 glioma cells cultured in the presence of serum or in a serum-free defined medium. Exponentially growing cells were seeded in multi-dish plates (5 x 10(4) cells/2 cm2 well) in DMEM supplemented with 5% fetal calf serum and were continuously exposed to 1 microgram/ml to 1,000 micrograms/ml suramin. Growth rate (determined 9 days after seeding) was reduced by 5%, 33%, 56% and 97%, respectively for suramin concentrations of 1, 10, 100 and 1000 micrograms/ml. Similar results were obtained in serum-free defined medium (DMEM/F12, 1:1, v:v, EGF 5 ng/ml, transferrin 5 micrograms/ml, selenium 10 ng/ml). Moreover, the concentration of suramin in the culture medium remained constant, demonstrating that the drug was not actively metabolized by the cells. Suramin also induced marked changes in cell morphology: the usual bipolar shape of C6 cells evolved toward a more differentiated appearance, with numerous cellular processes allowing a wide number of cell-cell contacts. In parallel, we monitored expression of an adhesion molecule (N-CAM) at both the mRNA and protein levels. Indirect immunofluoresence technique showed an important increase in cell surface N-CAM expression, starting from a dose of 10 micrograms/ml suramin, whereas total cellular content of N-CAM protein as well as its mRNA levels were unaffected. We also observed that the levels of expression of actin and N-CAM mRNAs decreased by a factor of two in cells maintained in defined medium. However, the relative ratio of N-CAM mRNA over actin mRNA was virtually unchanged following suramin treatment. Taken together, our results suggest that suramin (i) exerts a blocking effect of autocrine growth factors, (ii) interferes with the turn-over mechanisms of N-CAM expressed at the cell surface, either by impairing its endocytosis and/or the process of release of the N-CAM 120 isoform.

CNS neuronal cell line-derived factors regulate gliogenesis in neonatal rat brain cultures

Recent studies suggest that heterotypic cell-cell interactions influence gliogenesis in the developing rat central nervous system. CNS neuron-derived factors have been hypothesized to exist, and several have been identified and partially characterized which affect the number of oligodendrocytes in vitro. In order to study further the role of neurons in gliogenesis, we have used serum-free culture conditions, the B104 CNS neuronal cell line as a source of soluble factors, and dissociated neonatal rat brain cells as a source of glial cells. We have analyzed the response of the glial cells to serum-free B104 conditioned medium using morphological, immunocytochemical, autoradiographic, and enzymatic methods. Dose-dependent increases in the number of morphologically identified oligodendrocytes occur in response to this conditioned medium. Galactocerebroside (GalC) is a specific marker for oligodendrocytes, and the A2B5 antigen marks bipotential glial progenitor cells and their progeny: immature oligodendrocytes and type 2 astrocytes. In the presence of conditioned medium, the number of cells expressing GalC and/or A2B5 antigen increases over time when measured at 4, 8, and 12 days in vitro. A significantly weaker effect is seen if serum is also present. Since the vast majority of A2B5-positive cells in conditioned medium treated cultures lack glial fibrillary acidic protein (GFA), indicative of type 2 astrocytes, they represent glial progenitors and immature oligodendrocytes. Double immunostaining combined with autoradiography suggests that the latter cell types are the target cells for the oligodendrocyte-promoting activity. In addition, the conditioned medium markedly increases 2',3' cyclic nucleotide 3'-phosphodiesterase (an oligodendrocyte marker) and to a lesser extent enhances glutamine synthetase activity (an astrocyte marker). Type 1 astrocytes are also more morphologically differentiated in this condition, and their percentage is decreased simultaneously. Conditioned medium from other donor neural cells either has no activity or is much less effective than B104 conditioned medium. The active factors are soluble, sensitive to both trypsin and 100 degrees C treatment for 20 min, and appear to be 30-100 kilodaltons by stirred cell ultrafiltration. In summary, we have identified a potent source of growth-stimulating factors that produce increased numbers of glial progenitor cells and oligodendrocytes; the same conditioned medium also appears to inhibit type 1 astrocyte proliferation.

Effect of epidermal growth factor and insulin on DNA, RNA, and cytoskeletal protein labeling in primary rat astroglial cell cultures

The effect of epidermal growth factor (EGF) and insulin on DNA, RNA, and cytoskeletal protein labeling in primary rat astroglial cell cultures was investigated. Cultures were grown for 15-30 days in vitro in 10% fetal calf serum (FCS)-supplemented medium and then maintained in serum-free basal medium (DMEM) supplemented with fatty acid-free bovine serum albumin (BSA) for a starvation period of 24 hr before the addition of factors. The effect of factors was tested at different times (4, 10, 22, and 28 hr). At each time, [methyl-3H]thymidine or [5,6-3H]uridine was added to the control and treated cells; the incubation time after the addition of labeled precursors was 2 hr at 37 degrees C. The results obtained indicated that the addition of EGF or FCS significantly stimulated [methyl-3H]thymidine incorporation into DNA, reaching the maximum effect after 22 hr. EGF alone significantly stimulated [3H]uridine incorporation into RNA, and this effect was already maximum at 4 hr and remained constant up to 22 hr. The addition of insulin alone caused a slight increase in nucleic acid labeling for short times (4-10 hr). In contrast with EGF, no detectable stimulation of incorporation of labeled precursors after insulin treatment for 22 hr was observed. On the other hand, the addition of insulin in the presence of EGF induced an increase of the values observed with EGF alone on macromolecular synthesis at all the times studied. Furthermore, a decrease in cell number was observed in confluent cultures maintained for 1 week in medium containing DMEM + BSA in comparison to serum-supplemented (DMEM + BSA + FCS) cultures.

A chemically defined medium for the culture of mature oligodendrocytes

A new chemically defined medium consisting of equal parts of Dulbecco modified Eagle's and Ham's F-12 media supplemented with insulin, sodium selenite, putrescine, and D+ galactose, which allows the long-term survival of mature oligodendrocyte pure cultures, is described. Immunohistochemical staining has shown that over 90% of the cells become positive for myelin proteins shortly following subculture. Contaminating astrocytes (2%) do not survive in this medium. Biochemical data have indicated that these purified oligodendrocytes express 2'3'-cyclic nucleotide 3'-phosphohydrolase and UDP-galactose ceramide galactosyltransferase activities. Electron microscopical examination revealed that the oligodendrocytes were mostly of medium-dark type and appeared to be identical to cells cultured in serum-containing medium. The ability to maintain pure oligodendrocyte cultures in such a defined medium will allow investigations concerning exogenous and endogenous factors involved in oligodendrocyte metabolism.

Adrenocorticoid receptors in C6 glioma cells: effects on cell growth

Rat C6 glioma cells contain two receptors for adrenocorticoids--the predominant glucocorticoid receptor and low densities of the Type I corticosteroid (mineralocorticoid) receptor. Nanomolar concentrations of deoxycorticosterone, corticosterone and aldosterone, which fully occupy Type I receptors, produced a slight stimulatory effect on C6 cell growth in serum-free media. However, spironolactone, a Type I receptor antagonist, and pregnenolone, which does not bind to Type I receptors, had similar effects. Therefore, the slight growth stimulation produced by low steroid concentrations is not mediated by Type I or glucocorticoid receptors, but may be due to an effect on cell membrane properties or other receptor-independent action. Occupation of glucocorticoid receptors by higher concentrations of corticosteroids inhibited C6 cell growth.

Growth requirements in vitro of oligodendrocyte cell lines and neonatal rat brain oligodendrocytes

I have defined the basic requirements for the proliferation of cell lines expressing oligodendrocyte properties and for the survival of galactocerebroside-positive oligodendrocytes derived from neonatal rat brains. Conventional serum-containing medium can be replaced by 01 medium, a chemically defined medium supplemented with insulin, transferrin, sodium selenite, and biotin. Thyroid hormone is not required. When cells are plated directly into O1 medium, the substratum has to be modified by precoating with polylysine and adding fibronectin to the medium prior to the cells. Both cell lines and brain cells can be subcultured numerous times in O1 medium without initial culture in serum-containing medium. Brain cultures can be maintained in O1 medium for several months and contain a significantly higher percentage of mature oligodendrocytes, a lower number of astrocytes, and no fibroblasts as compared to cells maintained in serum-containing medium.

Primary retinal and cortical glial cell cultures: effects of medium and serum on attachment and growth

Glial cells of the retina are anatomically distinctive and are thought to contribute importantly to retinal electrophysiology. However, no adequate preparation exists for studying them in isolation, in vitro. This report provides guidelines for primary retinal glial cultures (RET) and compares basal tissue culture features with those for neocortical glia (CX) and the well-studied rat glial line, C6. Cell attachment and growth of RET, CX, and C6 are unique. These differences are explored by the use of specific media and sera. RET attachment, unlike that for CX or C6, was far more sensitive to medium than serum. RET cells attached least quickly, CX most quickly; 4 hr after plating 20% of RET remained unattached. RET growth was poor and relatively insensitive to medium. In contrast, growth of CX or C6 was medium dependent. Serum had substantial effects on the growth of all three glial lines. Pig, goat, horse, and dog sera were particularly effective, often comparing favorably to fetal calf serum. Medium or serum optimal for cell attachment, typically, was not optimal for growth and serum effects were more dramatic than those of medium. By all measures, CX and C6, both derived from brain, were more alike than were the two rabbit primaries, CX and RET. The data reveal substantial differences between presumably similar cells and indicate a need for an empirically based choice of both basal-salt media and serum to optimize specific aspects of cell development in culture.

Selenium deficiency in cultured adrenocortical cells: restoration of glutathione peroxidase and resistance to hydroperoxides on addition of selenium

Cultured bovine adrenocortical cells were previously shown to be functionally deficient in selenium and vitamin E when grown in medium supplemented with fetal bovine serum. In the present experiments, the lack of significant bioavailable amounts of selenium in the medium was demonstrated by the finding of only low levels of glutathione peroxidase in the cultured cells (0.008 U/mg protein compared with 0.045 U/mg protein in fresh adrenocortical tissue). When 20 nM selenium as selenite was added to the cultured adrenocortical cells, glutathione peroxidase activity increased continuously over 72 h, with a total increase of about eightfold over this period. Over the same time-course, the highest concentration of cumene hydroperoxide tolerated by the cells without cell death increased progressively from 10 microM to 50 microM. Addition of 1 microM alpha-tocopherol also increased the amount of cumene hydroperoxide tolerated to 50 microM. Cell death was measured by cloning efficiency after removal of cumene hydroperoxide. Addition of either selenium or alpha-tocopherol had little effect on the growth rate of the cells over six passages, even when residual vitamin E was removed from the serum by extraction with ether and residual low molecular weight selenium compounds were removed by dialysis. It is concluded that combined deficiency of selenium and vitamin E, at least in the presence of other components of fetal bovine serum, has little effect on the ability of the cells to survive under normal conditions, as evidenced by continued long-term proliferation. However, the low levels of glutathione peroxidase resulting from selenium deficiency cause an increase susceptibility to peroxide-mediated toxicity. The combined deficiency of selenium and vitamin E impairs the ability of cells to survive under adverse conditions, as well as altering mitochondrial functions, as previously demonstrated.

Culture of adult rabbit retinal glial cells: methods and cellular origin of explant outgrowth

Culture methods to propagate glial cells from the avascular adult rabbit neural retina are described. To determine the site(s) in the retina from which the cells originated, retina fragments were retrieved from culture at intervals after explanation and processed for light microscopy to localize surviving cells. Tritiated-thymidine radioautography was used to determine the time of onset and the localization of proliferating cells in the isolated retina and in the early culture outgrowth. Although some glial cells located just subjacent to the inner limiting membrane and at the interface between the ganglion cell and inner plexiform layers were activated to DNA synthesis in three to five days after explanation of the retina, the dominant cell type in the cultures derived from the retina appeared to be the Muller cell. Regionally, in isolated retina fragments, Muller cells became hyperplastic and formed tissue masses of proliferating cells surrounding photoreceptor remnants. These proliferating clusters could be retrieved from the culture medium after about 10 days in vitro and replated whereupon they attached to the culture substrate and gave rise to cellular outgrowths. The cells in the early explant outgrowth were heteromorphic, but passaged cultures contained a relatively homogeneous population of cells that exhibited a low maximal growth rate and senesced quickly in vitro.

beta-Endorphin: characterization of binding sites specific for the human hormone in human glioblastoma SF126 cells

The established human glioblastoma cell line SF126 was found to bind tritiated human beta-endorphin (beta h-EP) in a saturable fashion. From displacement studies, the ED50 was estimated to be about 2.5 nM. The Kd was estimated as 1.9 X 10(-9) M and Scatchard analysis showed a biphasic pattern with a predominant low-affinity component. Binding reached a maximum at about 90 min at 22 degrees C and was instantaneously reversible. Tritiated [D-Ala2,D-Leu5]enkephalin and tritiated dihydromorphine did not bind to the cells. Sodium at a concentration of 150 mM decreased the specific binding by 80%. The interaction with the cellular binding site appeared to be mediated by the COOH-terminal segment of beta h-EP, as beta h-EP-(6-31) retained a high potency for displacing tritiated beta h-EP, and beta h-EP-(1-27) has no activity. Camel beta-EP was only about 1% as active as the human hormone.

Culture of rat cerebral oligodendrocytes in a serum-free, chemically defined medium

Oligodendrocytes were isolated from the cerebra of young rats (5-10 days old) by trypsinization of the tissue followed by cell separation on Percoll gradients. The isolation was carried out in physiological, isotonic media. The cell yield was 2-4 X 10(6) cells per brain; the plating efficiency was greater than or equal to 70%. Isolated cells were seeded on poly-L-lysine-coated culture dishes and maintained in a serum-free, chemically defined medium for at least 30 days. After 10 days in culture 67 +/- 10% of the surviving cells were oligodendrocytes, as judged by immunocytochemical and morphological criteria, whereas most of the other cells reacted positively with antiserum against glial fibrillary acidic protein. The expression of typical oligodendrocyte markers (2':3'-cyclic-nucleotide 3'-phosphodiesterase, galactocerebrosides and myelin basic protein) was greatly enhanced under these serum-free conditions as compared with cultures in serum-containing medium. The antigenic markers (galactocerebrosides, myelin basic protein) were absent in the freshly isolated cells but could be detected after 3 days in culture by immunocytochemistry. The activity of 2':3'-cyclic-nucleotide 3'-phosphodiesterase increased from 75 nmol min-1 mg-1 protein on day 4 to 400 nmol min-1 mg-1 protein on day 14 in culture.

Growth from regenerating goldfish retinal cultures in the absence of serum or hormonal supplements: tissue extract effects

The minimal requirements for the regeneration of optic nerve fibers in vitro were established in a serum-free retinal explant preparation. This serum-free preparation was developed as a prerequisite for testing the growth-promoting activity of tissue extracts prepared from the primary target of regenerating fibers. Explants taken from goldfish retinas 14 days after a prior optic nerve crush were capable of long-term survival and regenerated neurite outgrowth without serum or hormonal supplements. Serum-free conditions for explant outgrowth required only a basic Leibovitz (L-15) media containing 0.6% methyl cellulose (MC). Explants were also capable of neurite outgrowth in L-15 media alone when culture dishes were preplated with MC. MC treatment permitted both the regeneration of neurites in serum-free L-15 and a significant increase in the rate and extent of neurite outgrowth when combined with 10% fetal calf serum (FCS). Explants grown in L-15 with both MC and FCS produced a 2.5-fold increase in the length of neurite outgrowth over MC alone and a 1.5-fold increase in the length of neurite outgrowth over FCS alone. MC activity which permitted minimal serum-free regeneration and optimal serum supplemented regeneration was determined to be substrate related. Retinas were dissociated to determine if ganglion cells, like the intact explant, were capable of survival and neurite regeneration in serum-free conditions. These cells survived and extended long neurites when grown in L-15 with FCS or with FCS and MC, but they did not survive in serum-free L-15 with MC. The minimal serum-free conditions for explant survival and neurite regeneration were used as a model system to test the growth-promoting activity of crude tissue extracts prepared from the goldfish brain. Extracts prepared from the primary target region, the optic tectum, stimulated a significant 2.5-fold increase in the length of regenerating neurites. The optic tectal extract (OTex) stimulated outgrowth with significantly high specific activity when compared with extracts of identical protein concentrations prepared from the cerebellum (Cex). At a minimal protein concentration of 150 micrograms/ml, the OTex stimulated a 1.5-fold increase in neurite outgrowth above Cex. These results indicated that a serum-free culture preparation had been established for optic nerve regeneration. This culture system has proven to be an extremely sensitive bioassay model without the masking effect of a serum supplement. Serum-free cultures may be used in further studies to determine the role neurotrophic factors may play in a widely used model of successful central nervous system (CNS) regeneration.