BMP-7 and excess glutamate: opposing effects on dendrite growth from cerebral cortical neurons in vitro

Glutamate is an important regulator of dendrite development. During cerebral ischemia, however, there is massive release of glutamate reaching millimolar concentrations in the extracellular space. An early consequence of this excess glutamate is reduced dendrite growth. Bone morphogenetic protein-7 (BMP-7) a member of the transforming growth factor-beta (TGF-beta) superfamily has been demonstrated to enhance dendrite output from cerebral cortical and hippocampal neurons in vitro. However, it is not known whether BMP-7can prevent the reduced dendrite growth associated with excess glutamate or enhance dendrite growth after glutamate exposure. Therefore we quantified axon and primary, secondary, and total dendrite growth from embryonic mouse cortical neurons (E18) grown at low density in vitro in a chemically defined medium and exposed to glutamate (1 or 2 mM) for 48 h. Morphology and double immunolabeling (MAP2, NF-H) were used to identify cortical dendrites and axons after 3 DIV. In these short-term cultures, glutamate did not influence neuron survival. The addition of glutamate to cortical neurons, however, significantly attenuated dendrite output. This effect was mimicked by the addition of NMDA but not AMPA agonists and inhibited by the specific NMDA receptor antagonist MK-801. The reduction in dendrite growth mediated by excess glutamate was ameliorated by the administration of 30 or 100 ng/ml of BMP-7. In addition, when administered in a delayed fashion between 1 and 24 h after the initial glutamate exposure, BMP-7 was able to enhance dendrite growth, including primary dendrite number, primary dendrite length, and secondary dendritic branching. These findings demonstrate that BMP-7 can ameliorate reduced dendrite growth from cerebral cortical neurons associated with excess glutamate in vitro and are important because they may help explain why BMP-7 administration is associated with enhanced functional recovery in models of cerebral ischemia.     

OP-1 enhances dendritic growth from cerebral cortical neurons in vitro

Osteogenic protein-1 (OP-1), a member of the transforming growth factor-beta (TGF-beta) superfamily, has been demonstrated to stimulate dendrite growth from sympathetic neurons in culture. However, it is not known whether OP-1 affects dendrite growth from central nervous system neurons. Therefore we quantified axon and primary, secondary, and total dendritic growth from embryonic mouse cortical neurons (E 18) grown in vitro in a chemically defined medium. Morphology and double immunolabeling (MAP2, NF-H) were used to identify cortical dendrites and axons after 3 days in vitro. Cell morphology, neuron survival, and axon length were similar under all experimental conditions. The number of primary dendrites also was similar; however, the length of primary dendrites and the length and number of secondary dendrites were significantly increased by the addition of OP-1 to the culture medium. This increase in dendrite growth was dose-dependent; maximal dendritic growth was observed after the addition of 30-100 ng/ml of OP-1 to the culture medium. Specific support of dendrite growth was not observed when neurons were exposed to other members of the TGF-beta superfamily. These findings demonstrate that OP-1 selectively increases dendrite growth from cerebral cortical neurons in vitro.     

Glutamate receptor agonist kainate enhances primary dendrite number and length from immature mouse cortical neurons in vitro

Glutamate is an important regulator of dendrite development that may inhibit, (during ischemic injury), or facilitate (during early development) dendrite growth. Previous studies have reported mainly on the N-methyl-D-aspartate (NMDA) receptor-mediated dendrite growth-promoting effect of glutamate. In this study, we examined how the non-NMDA receptor agonist kainate influenced dendrite growth. E18 mouse cortical neurons were grown for 3 days in vitro and immunolabeled with anti-microtubule-associated protein 2 (MAP2) and anti-neurofilament (NF-H), to identify dendrites and axons, respectively. Exposure of cortical neurons to kainate increased dendrite growth without affecting neuron survival. This effect was dose-dependent, reversible and blocked by the alpha-amino-3-hydroxy-5-methyl-4-isoxazoleproprionate (AMPA)/kainate receptor antagonist NBQX and the low-affinity kainate receptor antagonist NS-102, but not by the AMPA receptor antagonist CFM-2. In addition, the NMDA receptor antagonist MK-801 had no effect on kainate-induced dendrite growth. Immunolabeling and Western blot analysis of kainate receptors using antibodies against the GluR6 and KA2 subunits, demonstrated that the immature cortical neurons used in this study express kainate receptor proteins. These results suggest that kainate-induced non-NMDA receptor activation promotes dendrite growth, and in particular primary dendrite number and length, from immature cortical neurons in vitro, and that kainate receptors may be directly involved in this process. Furthermore, these data support the possibility that like NMDA receptors, kainate receptor activation may also contribute to early neurite growth from cortical neurons in vitro.     

Reduced dendrite growth and altered glutamic acid decarboxylase (GAD) 65- and 67-kDa isoform protein expression from mouse cortical GABAergic neurons following excitotoxic injury in vitro

The vulnerability of brain cells to neurologic insults varies greatly, depending on their neuronal subpopulation. However, cells surviving pathological insults such as ischemia or brain trauma may undergo structural changes, e.g., altered process growth, that could compromise brain function. In this study, we examined the effect of glutamate excitotoxicity on dendrite growth from surviving cortical GABAergic neurons in vitro. Glutamate exposure did not affect GABAergic neuron viability, however, it significantly reduced dendrite growth from GABAergic neurons. This effect was blocked by the AMPA receptor antagonists NBQX and CFM-2, and mimicked by AMPA, but not NMDA. Glutamate excitotoxicity also caused an NMDA receptor-mediated decrease in the GABA synthesizing enzyme glutamic acid decarboxylase (GAD65/67) immunoreactivity from GABAergic neurons, measured using immunocytochemical and Western blot techniques. GAD is necessary for GABA synthesis; however, reduction of GABA by 3-mercaptopropionic acid (3-MPA), which inhibits GABA synthesis, did not alter dendrite growth. These results suggest that GABAergic cortical neurons are relatively resistant to excitotoxic-induced cell death, but they can display morphological and biochemical alterations which may impair their function.     

Effect of excess extracellular glutamate on dendrite growth from cerebral cortical neurons at 3 days in vitro: Involvement of NMDA receptors

Glutamate is an important regulator of dendrite development; however, during cerebral ischemia, massive glutamate release can lead to neurodegeneration and death. An early consequence of glutamate excitotoxicity is dendrite injury, which often precedes cell death. We examined the effect of glutamate on dendrite growth from embryonic day 18 (E18) mouse cortical neurons grown for 3 days in vitro (DIV) and immunolabeled with anti-microtubule-associated protein (MAP)2 and anti-neurofilament (NF)-H, to identify dendrites and axons, respectively. Cortical neurons exposed to excess extracellular glutamate (100 microM) displayed reduced dendrite growth, which occurred in the absence of cell death. This effect was mimicked by the ionotropic glutamate receptor agonist N-methyl-D-aspartate (NMDA) and blocked by the ionotropic glutamate receptor antagonist kynurenic acid and the NMDA receptor-specific antagonist MK-801. The non-NMDA receptor agonist AMPA, however, did not affect process growth. Neither NMDA nor AMPA influenced neuron survival. Immunolabeling and Western blot analysis of NMDA receptors using antibodies against the NR1 subunit, demonstrated that immature cortical neurons used in this study, express NMDA receptors. These results suggest that excess glutamate decreases dendrite growth through a mechanism resulting from NMDA receptor subclass activation. Furthermore, these data support the possibility that excess glutamate activation of NMDA receptors mediate both cell death in mature neurons and the inhibitory effect of excess glutamate on dendrite growth in immature neurons or in the absence of cell death.     

Astrocytes mediate cerebral cortical neuronal axon and dendrite growth, in part, by release of fibroblast growth factor.

Astrocytes occupy a central role in central nervous system (CNS) function. In particular astyrocytes can support neurite growth, in part, by release of diffusable factors. We therefore performed biochemical analysis of astrocyte conditioned medium to examine possible mechanisms of astrocyte mediated axon and dendrite growth in the mammalian CNS. Culture medium was conditioned on purified astrocyte monolayers derived from P3 rat cerebral cortex or on fibroblasts. Conditioned medium (CM) was subject to protein denaturation, molecular weight fractionation, and heparin affinity chromatography. E18 mouse cerebral cortical neurons were then cultured in the various media or directly on astrocyte monolayers and axon and dendrite growth from 50 neurons in each condition quantified after 3 DIV using double-labeled immunohistochemical techniques. Axon and dendrite growth was supported by astrocyte CM and both were significantly greater than process growth from neurons incubated in fibroblast CM. Protein denaturation significantly reduced astrocyte CM support of axon and dendrite growth. Following ultrafiltration and dialysis dendrite and axon growth was observed in the molecular weight fraction between 10 and 100 kDa. Axon growth also was observed in the CM molecular weight fraction greater than 100 kDa. Conditioned medium was eluted on a heparin column; when the bound fragment was reconstituted in chemically defined medium extensive dendrite and axon growth was observed. Since fibroblast growth factor (FGF) has these biochemical characteristics we added anti-bFGF neutralizing antibodies to astrocyte monolayers or CM; this significantly reduced astrocyte support of process growth. By contrast, the addition of heparin, which helps activate FGF receptors, to astrocyte CM further enhanced process growth. Western blot analysis confirmed that bFGF was present in astrocyte CM. We then examined axon and dendrite growth from cortical neurons after the addition of various growth factors to chemically defined medium. Axon and dendrite growth, similar to that found in astrocyte CM was observed after the addition of bFGF or aFGF. Astrocyte support of cerebral cortical neuron axon and dendrite growth in vitro may be explained, in part, by FGF release.     

Cytokine‐induced changes in the ability of astrocytes to support migration of oligodendrocyte precursors and axon growth

Repair of demyelination in the CNS requires that oligodendrocyte precursors (OPs) migrate, divide and then myelinate. Repair of axon damage requires axonal regeneration. Limited remyelination and axon regeneration occurs soon after injury, but usually ceases in a few days. In vivo and in vitro experiments have shown that astrocytic environments are not very permissive for migration of OPs or for axonal re-growth. Yet remyelination and axon sprouting early after injury occurs in association with astrocytes, while later astrocytes can exclude remyelination and prevent axon regeneration. A large and changing cast of cytokines are released following CNS injury, so we investigated whether some of these alone or in combination can affect the ability of astrocytes to support migration of OPs and neuritic outgrowth. Interleukin (IL) 1α, tumour necrosis factor α, transforming growth factor (TGF) β, basic fibroblast growth factor (bFGF), platelet-derived growth factor and epidermal growth factor alone exerted little or no effect on migration of OPs on astrocytes, whereas interferon (IFN) γ was inhibitory. The combination of IL-1α + bFGF was found to be pro-migratory, and this effect could be neutralized by TGFβ. We also examined neuritic outgrowth from dorsal root ganglion explants in three-dimensional astrocyte cultures treated with cytokines and found that IL-1α + bFGF greatly increased axon outgrowth and that this effect could be blocked by TGFβ and IFNγ. All these effects were absent or much smaller when OP migration or axon growth was tested on laminin, so the main effect of the cytokines was via astrocytes. The cytokine effects did not correlate with expression on astrocytes of laminin, fibronectin, tenascin, chondroitin sulphate proteoglycan, N-cadherin, polysialyated NCAM (PSA-NCAM), tissue plasminogen activator (tPA) or urokinase (uPA).     

Oxidative stress in rat cortical neurons and astrocytes induced by paraquat in vitro

Oxidative stress has been discussed as crucial mechanism of neuronal cell death in the adult brain. However, it was not clear until now whether neurons are more sensitive to oxidative stress than the other cells in the brain, e.g. astrocytes. Therefore both cell types were exposed to oxidative stress provoked by the redox-cycling compound paraquat. Cortical neurons were found to be more sensitive towards paraquat toxicity than astrocytes as shown by MTT and Neutral Red assay, two different cytotoxicity assays. Mitochondrial functions were determined by the mitochondrial membrane potential and intracellular ATP concentrations. Again cortical neurons were more severely impaired (by paraquat than astrocytes). The production of reactive oxygen species after paraquat exposure was much higher in cortical neurons than in astrocytes and correlated with a higher depletion of GSH (intracellular glutathion). Lipid peroxidation could be shown in astrocytes via the breakdown product malondialdehyde (MDA) whereas in cortical neurons 4-hydroxynonenal (4-HNE) was detected as this endpoint. If and how oxidative stress influences the antioxidant defense was determined via changes in the expression of antioxidant enzymes. Paraquat exposure lead to a 2-3 fold increase of catalase, MnSOD and CuZnSOD mRNA expression in astrocytes. In contrast to astrocytes, in cortical neurons catalase and MnSOD mRNA levels were only marginally elevated above 1.5-fold by treatment with paraquat. Expression levels of glutathione peroxidase (GPx) mRNA were the only one that were not changed in both cell types after paraquat exposure. It is concluded that the more marked increase in expression levels of antioxidant enzymes may render astrocytes more resistant to oxidative stress than neuronal cells.     

NGF-dependent axon growth and regeneration are altered in sympathetic neurons of dystrophic mdx mice

Duchenne muscular dystrophy (DMD) is a lethal disease, determined by lack of dystrophin (Dp427), a muscular cytoskeletal protein also expressed by selected neuronal populations. Consequently, besides muscular wasting, both human patients and DMD animal models suffer several neural disorders. In previous studies on the superior cervical ganglion (SCG) of wild type and dystrophic mdx mice (Lombardi et al. 2008), we hypothesized that Dp427 could play some role in NGF-dependent axonal growth, both during development and adulthood. To address this issue, we first analyzed axon regeneration potentials of SCG neurons of both genotypes after axotomy in vivo. While noradrenergic innervation of mdx mouse submandibular gland, main source of nerve growth factor (NGF), recovered similarly to wild type, iris innervation (muscular target) never did. We, therefore, evaluated whether dystrophic SCG neurons were poorly responsive to NGF, especially at low concentration. Following in vitro axotomy in the presence of either 10 or 50 ng/ml NGF, the number of regenerated axons in mdx mouse neuron cultures was indeed reduced, compared to wild type, at the lower concentration. Neurite growth parameters (i.e. number, length), growth cone dynamics and NGF/TrkA receptor signaling in differentiating neurons (not injured) were also significantly reduced when cultured with 10 ng/ml NGF, but also with higher NGF concentrations. In conclusion, we propose a role for Dp427 in NGF-dependent cytoskeletal dynamics associated to growth cone advancement, possibly through indirect stabilization of TrkA receptors. Considering NGF activity in nervous system development/remodeling, this aspect could concur in some of the described DMD-associated neural dysfunctions.     

Quantitative studies on proliferative changes of reactive astrocytes in mouse cerebral cortex

Cell number and proliferation of reactive astrocytes were studied quantitatively in the stabbed cerebral cortex of adult mice, using immunohistochemistry for glial fibrillary acidic protein (GFAP) and [3H]thymidine autoradiography. GFAP-positive astrocytes increased in cell number gradually from 24 to 96 h after stabbing, and their immunoreactivity became intense. The maximum number of GFAP-positive cells was about 4.5 times normal in the layers II-VI of the cortex, whereas it was only 1.5 times normal in the layer I (molecular layer). In contrast to the gradual increase in cell number, no GFAP-positive astrocytes were labeled with [3H]thymidine prior to 48 h after stabbing, in either the layer I or the layers II-VI. Then 3-5% of them were labeled at 72 and 96 h, but very few again after 6 days. By injecting [3H]thymidine successively for 6 days after stabbing, only 17% of GFAP-positive astrocytes of the layer I or the layers II-VI were labeled. These results reveal that, in the cortical layers II-VI, many GFAP-negative source cells initially express much more GFAP-antigen without proliferation and change into GFAP-positive reactive astrocytes. Proliferation of reactive astrocytes is not the major factor for the marked increase in number of them. The cortical layer I would have few GFAP-negative source cells for reactive astrocytes. These source cells may be protoplasmic astrocytes.     

Primates exposed to cocaine in utero display reduced density and number of cerebral cortical neurons.

This study examined the effects of cocaine use during the second trimester of pregnancy on cerebral neocortical volume and density, and total number of neocortical neurons and glia in offspring. We also evaluated the extent of postnatal recovery of cytoarchitectural abnormalities previously observed in the neocortex of two-month-old primates born from cocaine-treated mothers (Lidow [1995] Synapse 21:332-334). Pregnant monkeys received cocaine orally (20 mg/kg/day) from the 40th to 102nd days of pregnancy (embryonic day [E]40-E102). On E64 and E65, the animals were injected with [(3)H]thymidine. Cerebral hemispheres of the offspring were examined at three years of age. We found a reduction in the neocortical volume and density and total number of neocortical neurons. The observed reduction in neuronal number within the neocortex was not accounted for by the increase in the number of neurons in the white matter of cocaine-exposed animals, because the number of these "extra" neurons was equal to only half that of missing neurons. We detected no significant changes in the number of neocortical glia. The cytoarchitectural abnormalities in the neocortex of prenatally cocaine-exposed three-year-old monkeys closely resembled previously described neocortical abnormalities in similarly exposed two-month-old animals: the neocortex lacked a discernible lamination; the majority of the cells labeled by [(3)H]thymidine injected during neocortical neurogenesis did not reach their proper position within the cortical plate. Therefore, postnatal maturation is not associated with significant improvement in neocortical organization in primates prenatally exposed to cocaine. There was, however, a postnatal recovery of low glial fibrillary acidic protein (GFAP) immunoreactivity previously observed in 2-month-old cocaine-exposed animals.     

重组人Semaphorin3A诱导体外培养的大鼠大脑皮层神经元轴突损伤的实验研究

目的探讨外源性Semaphorin3A(Sema3A)对于体外培养神经元轴突生长及神经元生长活性的影响。方法体外培养新生SD大鼠皮质神经元,随机分为正常对照组和Sema3A不同浓度处理组,倒置相差显微镜及微管相关蛋白-2(MAP-2)荧光染色分别观察生长锥及轴突形态学变化;噻唑蓝(MTT)法检测神经元存活率。结果 Sema3A(0.5mg/ml)处理可诱发神经元生长锥崩解;与正常对照组相比,Sema3A(5mg/ml)处理组轴突平均长度缩短(P0.001);经不同浓度Sema3A处理后,神经元存活率呈剂量依赖性下降,其中浓度范围为0.5~5.0mg/ml的处理组与正常组比较均有统计学意义(P0.001)。结论 Sema3A在体外可发挥明显的促生长锥崩解及抑制轴突生长的作用,并且具有一定的神经元毒性。     

Differential expression of glutamate dehydrogenase in cultured neurons and astrocytes from mouse cerebellum and cerebral cortex.

Glutamate dehydrogenase (GDH) specific activities, kinetic properties and allosteric regulation were studied in extracts from cultured neurons and astrocytes prepared from mouse cerebral cortex and cerebellum. Considerable differences were observed in the specific activity of the enzyme among the different cell types with astrocytes expressing the highest GDH activity. This may reflect the functional importance of these cells in glutamate uptake and metabolism. Among the neurons, the glutamatergic cerebellar granule cells showed a GDH specific activity that was 60% higher (P < 0.01) than that of the GABAergic cerebral cortical neurons. Also, the K(m) for ammonia was 1.7-fold higher in the cortical neurons than in the other cell types. These findings may reflect a particular need for the glutamatergic granule cells to synthesize glutamate via the GDH pathway. No differences were observed among the different cell types with regard to the allosteric properties of GDH expressed by these cells.     

The SUMO protease Verloren regulates dendrite and axon targeting in olfactory projection neurons

Sumoylation is a post-translational modification regulating numerous biological processes. Small ubiquitin-like modifier (SUMO) proteases are required for the maturation and deconjugation of SUMO proteins, thereby either promoting or reverting sumoylation to modify protein function. Here, we show a novel role for a predicted SUMO protease, Verloren (Velo), during projection neuron (PN) target selection in the Drosophila olfactory system. PNs target their dendrites to specific glomeruli within the antennal lobe (AL) and their axons stereotypically into higher brain centers. We uncovered mutations in velo that disrupt PN targeting specificity. PN dendrites that normally target to a particular dorsolateral glomerulus instead mistarget to incorrect glomeruli within the AL or to brain regions outside the AL. velo mutant axons also display defects in arborization. These phenotypes are rescued by postmitotic expression of Velo in PNs but not by a catalytic domain mutant of Velo. Two other SUMO proteases, DmUlp1 and CG12717, can partially compensate for the function of Velo in PN dendrite targeting. Additionally, mutations in SUMO and lesswright (which encodes a SUMO conjugating enzyme) similarly disrupt PN targeting, confirming that sumoylation is required for neuronal target selection. Finally, genetic interaction studies suggest that Velo acts in SUMO deconjugation rather than in maturation. Our study provides the first in vivo evidence for a specific role of a SUMO protease during neuronal target selection that can be dissociated from its functions in neuronal proliferation and survival.     

Birth-date-dependent segregation of the mouse cerebral cortical neurons in reaggregation cultures

Cerebral cortical neurons form a six-layered structure in which their position depends on their birth date. This developmental process requires the presence of Reelin, which is secreted by Cajal-Retzius cells in the cortical marginal zone (MZ). However, it is still unclear whether the migration from the ventricular zone (VZ) to beneath the MZ is essential for the neurons to segregate into layers. Previous transplantation studies of ferret cerebral cortical neurons suggested that their ultimate laminar fate is, at least to some extent, determined in the VZ but it is unknown how 'laminar fate' eventually positions cells in a specific layer. To explore the segregation properties of mouse cortical cells that have not yet arrived beneath the MZ, embryonic day (E)16 VZ and intermediate zone (IMZ) cells were dissociated and allowed to reaggregate for 1-4 days in vitro. The results suggested that the migrating neurons in the IMZ at E16 preferentially located near the centre of the aggregates, more than did the proliferative cells from the VZ. The birth-date labelling followed by the dissociation-reaggregation culture suggested that the segregation properties of the E16 IMZ was characteristic of the E14-born cells, which were migrating in the IMZ at E16, but they were not general properties of migrating IMZ cells. This birth-date-dependent segregation mechanism was also observed in the Reelin signalling-deficient yotari cells. These findings suggest that cortical neurons acquire a birth-date-dependent segregation mechanism before their somas reach the MZ.     

Lithium-potassium interaction in acutely treated cortical neurons and astrocytes

Pure mouse primary cultures of cortical astrocytes and of cortical neurons were exposed to 1 mM Li⁺, i.e., a therapeutically relevant concentration. The ⁴²K uptake rates of neurons were not influenced, whereas those of astrocytes showed an 11% inhibition (P < 0.01). Internal loading with Li did not change the K⁺ uptake rates in either cell type. Neurons, which had been exposed for 5 min to yeratridine, a situation which mimics neuronal activity, showed also no change in K uptake rate when Li was present during this time. Na⁺ -K⁺ ATPase activity from cell homogenates was not changed in neuronal preparations by exposure to Li⁺, but astrocytic preparations appeared to show a slight increase by 14%. These experiments point out that the Li effects on ion distribution of the brain which have been described in the literature, are due to partly impairment of astrocytic K⁺ uptake. The mechanism of action, underlying the Li⁺ effect is probably a competition with K⁺for transport sites at the external site of the Na⁺ -K⁺ ATPase. This leads to a decrease of K uptake, but an enhancement of ATPase activity in the presence of Li⁺.     

Effects of early ethanol exposure on dendrite growth of cortical pyramidal neurons: inferences from a computational model

A computational model has been used to infer rules governing dendritic growth of layer 2/3 associative pyramidal neurons in a rat model of foetal alcohol syndrome. Basal dendrites were studied in adult rats exposed to ethanol during the first postnatal week. Results suggest that ethanol exposure during early postnatal life affects mainly the branching of dendrites rather than their elongation.     

Incidence of visual cortical neurons which have axon collaterals projecting to both cerebral hemispheres during prenatal primate development

Fast blue was injected massively in extrastriate cortex of one hemisphere Diamidino yellow in area 17 of the other hemisphere, in adult and prenatal cynomolgus monkeys. After a suitable survival period the brains were processed for fluorescent dyes. Counts were made of the total number of labeled neurons and of those neurons which were labeled by both dyes and which project therefore to both hemispheres by means of bifurcating axon collaterals. At 122 and 135 days after conception (E122 and E135), shortly after cortico-cortical pathways are established, double-labeled neurons constituted 0.45% and 0.46% of the total population of labeled neurons in area V2. In V2 in the adult the range of values of double-labeled neurons was 0.03-0.08%.     

Aluminum pretreatment impairs the ability of astrocytes to protect neurons from glutamate mediated toxicity

A number of laboratories have shown that astrocytes protect neurons from glutamate excitotoxicity. The experiments described in this paper were designed to address the question whether prior exposure of astrocytes to aluminum (in the form of aluminum citrate) interfered with the ability of astrocytes to protect neurons from glutamate excitotoxicity. Our culture paradigm consisted of highly enriched cultures of neurons and astrocytes grown on separate coverslips; this design enables one to subject either the neurons or the astrocytes to specific treatments and recombine the cells into the same petri dish simply by moving ceverslips from dish to dish. We have confirmed findings of other laboratories that astrocytes could protect from glutamate-induced death when glutamate (100 μM) is added to the culture medium. We have also demonstrated that prior treatment of astrocytes with 100 μM aluminum citrate impairs this ability of astrocytes to promote neuronal survival. No differences, however, were observed in the ability of control and aluminum-treated astrocytes to take up glutamate. These findings suggest that aluminum may cause astrocytes to: (i) secrete a factor that makes neurons more susceptible to glutamate-induced toxicity; (ii) secrete a neuronotoxic factor in the presence of glutamate; or (iii) reduce secretion of a factor that protects neurons from glutamate excitotoxicity.