Mitochondrial dysfunction and glutamate excitotoxicity studied in primary neuronal cultures

Primary dissociated neuronal cultures have been intensively exploited for the past 15 years as model systems to investigate excitotoxic neuronal degeneration. Even this simplified system contains a complex web of interactions between calcium homeostasis, ATP production and the generation and detoxification of reactive oxygen species. There is increasing realization that the mitochondrion occupies the center stage in these processes. This review covers the normal bioenergetics of the cultured neuron, the ways in which mitochondrial dysfunction impacts upon the ability of the neuron to withstand excitotoxic stress, the nature of the stresses imposed by NMDA receptor activation and possible molecular mechanisms of excitotoxic cell death.     

Glutamate receptor subunit expression in primary neuronal and secondary glial cultures

We report on the expression of ionotropic glutamate receptor subunits in primary neuronal cultures from rat cortex, hippocampus and cerebellum and of metabotropic glutamate (mGlu) receptor subtypes in these neuronal cultures as well as in cortical astroglial cultures. We found that the NMDA receptor (NR) subunits NR1, NR2A and NR2B were expressed in all three cultures. Each of the three cultures showed also expression of the four AMPA receptor subunits. Although RT-PCR detected mRNA of all kainate (KA) subunits in the three cultures, western blot showed only expression of Glu6 and KA2 receptor subunits. The expression analysis of mGlu receptors indicated the presence of all mGlu receptor subtype mRNAs in the three neuronal cultures, except for mGlu2 receptor mRNA, which was not detected in the cortical and cerebellar culture. mGlu1a/alpha, -2/3 and -5 receptor proteins were present in all three cultures, whereas mGlu4a and mGlu8a receptor proteins were not detected. Astroglial cultures were grown in either serum-containing or chemically defined medium. Only mGlu5 receptor protein was found in astroglial cultures grown in serum-containing medium. When astrocytes were cultured in chemically defined medium, mGlu3, -5 and -8 receptor mRNAs were detected, but at the protein level, still only mGlu5 receptor was found.     

Properties of single chloride channels in primary neuronal cultures of Drosophila

An anion channel from Drosophila neurons had the selectivity sequence: NO3- (1.97) greater than Br- (1.12) approximately equal to I- (1.03) approximately equal to Cl- (1) greater than F- (0.32) much greater than glutamate (less than 0.02) as estimated by the permeability ratio based on the reversal potential measurement. 4-Acetoamido-4'-isothiocyanostilbene-2,2'-disulfonic acid applied internally at 10 microM blocked the channel. We suggest that the chloride channel identified here may provide a pathway for Cl- in the resting membrane of Drosophila neurons.     

Intercellular junctions and transfer of small molecules in primary vascular endothelial cultures

The ultrastructure of gap and tight junctions and the cell-to-cell transfer of small molecules were studied in primary cultures and freshly isolated sheets of endothelial cells from calf aortae and umbilical veins. In thin sections and in freeze-fracture replicas, the gap and tight junctions in the freshly isolated cells from both sources appeared similar to those found in the intimal endothelium. Most of the interfaces in replicas had complex arrays of multiple gap junctions either intercalated within tight junction networks or interconnected by linear particle strands. The particle density in the center of most gap junctions was noticeably reduced. In confluent monolayers, after 3-5 days in culture, gap and tight junctions were present, although reduced in complexity and apparent extent. Despite the relative simplicity of the junctions, the cell-to-cell transfer of potential changes, dye (Lucifer Yellow CH), and nucleotides was readily detectable in cultures of both endothelial cell types. The extent and rapidity of dye transfer in culture was only slightly less than that in sheets of freshly isolated cells, perhaps reflecting a reduced gap junctional area combined with an increase in cell size in vitro.     

Comparison of prostanoid forming capacity of neuronal and astroglial cells in primary cultures

Prostaglandin (PG) and thromboxane (TX) biosynthesis in primary neuronal and astroglial cell cultures was studied. Cultures obtained from fetal (15–16 days old) and neonatal rat brain hemispheres were characterized by chemical and immunocytochemical staining techniques as predominantly neurons or mature and immature astrocytes, respectively. Six-day old neuronal cell cultures grown in the presence of cytosine arabinoside (2 μM) from the day 3 onwards were contaminated up to 10% with glioblasts. In astroglial cultures up to 3% of the cells were postively stained with a marker for oligodendroglial cells. Fibroblast contamination was below 1% in both cultures. Prostanoid formation (measured by specific radioimmunoassays) in 6-day old neuronal cell cultures was low (sum of the amount of PGs and TX formed: 1.16 ± 0.17 (ng/mg protein/15 min) as compared to 14-day old cultured astroglial cells: 21.27 ± 2.53 (ng/mg protein/15 min). Also the pattern of prostanoids formed was different in neuronal (PGD₂ ? PGF_2α > TXB₂ ? PGE₂) and astroglial cells (PGD₂ > TXB₂ ? PGF_2α ? PGE₂ ? 6-ketoPGF_1α). Preincubation with arachidonic acid (1 μg/ml) did not affect prostanoid formation in both cultures, whereas it was stimulated 4–6-fold by addition of the calcium ionophore A23187 (1 μM). These results, although found on cultured neuronal and glial cells of different stages of development, support the view that astroglial cells might play a crucial role in brain prostanoid synthesis.     

Computer analysis of organelle translocation in primary neuronal cultures and continuous cell lines

Organelle translocation in a number of cell types in tissue culture as seen by high-resolution Zeiss-Nomarski differential interference contrast optics was filmed and analyzed by computer. Principal cell types studied included primary chick spinal cord, chick dorsal root ganglion, ratbrain, and various clones of continuous cell lines. Organelle translocations in all cell types studied exhibited frequent, large changes in velocity during any one translocation. The appearance of particles as seen with Nomarski optics was correlated with their fine structures in one dorsal root ganglion neurite by fixing the cell as it was being filmed and obtaining electron micrographs of the region filmed. This revealed the identity of several organelles as well as the presence of abundant neurotubules but no neurofilaments. Primary cell cultures exhibited more high-velocity organelle movements than continuous cell lines. The net progress of an organelle in a given direction was greater in primary neuronal cells than in fibroblasts or continuous cell lines. These findings are correlated with the literature on organelle translocation and axoplasmic transport.     

Adrenergic and calcium-mediated subcellular redistribution of protein kinase C in primary neuronal cultures

Incubation of primary neuronal cultures prepared from the brains of neonatal rats with 50 microM epinephrine resulted in the transient redistribution of protein kinase C from the cytosol to the particulate fraction. This effect occurred after 1 and 5 min of incubation and resulted in a decrease in cytosolic protein kinase C activity with a corresponding increase in particulate protein kinase C of approximately 30% and 15%, respectively. The epinephrine-stimulated translocation of protein kinase C was blocked by 1 microM prazosin indicating the involvement of alpha 1-adrenergic receptors. Further, inclusion of 0.1 microM Ca2+ in the homogenization buffer was found to significantly enhance the binding of protein kinase C to cellular membranes prepared from neuronal cultures. These results indicate that alpha 1-adrenergic receptors in neuronal brain cell cultures are linked to the activation of protein kinase C and that the mobilization of Ca2+ may enhance this effect.     

Differences in the release ofl-glutamate andd-aspartate from primary neuronal chick cultures

Primary neuronal cultures were made from eight-day-old embryonic chick telencephalon. Ten-day-old cultures were used to study the release ofd-[³H]aspartate andl-[³H]glutamate. Thed-[³H]aspartate release was stimulated by increasing potassium concentrations, but it was not calcium dependent. In contrast, the potassium dependentl-[³H]glutamate release was calcium dependent, and furthermorel-[³H]glutamate release was optimal at potassium concentrations<30 mM. The inhibitors of glutamate uptake, dihydrokainate and 1-aminocyclobutane-trans-1,3-dicarboxylic acid (CACB), also referred to as cis-1-aminocyclobutane-1,3-dicarboxylate, were used in the release experiments. Dihydrokainate had no effect on aspartate release, whereas CACB increased both the basal efflux ofd-[³H]aspartate and the potassium evoked release. CACB had no effect on the potassium stimulatedl-glutamate release. We believe thatl-glutamate is released mainly by a vesicular mechanism from the presumably glutamatergic neurons present in our culture.d-aspartate release observed by us, could be mediated by a transporter protein. The cellular origin of this release remains to be assessed.     

Exacerbation of copper toxicity in primary neuronal cultures depleted of cellular glutathione

Perturbations to glutathione (GSH) metabolism may play an important role in neurodegenerative disorders such as Alzheimer's, Parkinson's, and prion diseases. A primary function of GSH is to prevent the toxic interaction between free radicals and reactive transition metals such as copper (Cu). Due to the potential role of Cu in neurodegeneration, we examined the effect of GSH depletion on Cu toxicity in murine primary neuronal cultures. Depletion of cellular GSH with L-buthionine-[S,R]-sulfoximine resulted in a dramatic potentiation of Cu toxicity in neurons without effect on iron (Fe) toxicity. Similarly, inhibition of glutathione reductase (GR) activity with 1,3-bis(2-chloroethyl)-1-nitrosurea also increased Cu toxicity in neurons. To determine if the Alzheimer's amyloid-beta (Abeta) peptide can affect neuronal resistance to transition metal toxicity, we exposed cultures to nontoxic concentrations of Abeta25-35 in the presence or absence of Cu or Fe. Abeta25-35 pretreatment was found to deplete neuronal GSH and increase GR activity, confirming the ability of Abeta to perturb neuronal GSH homeostasis. Abeta25-35 pretreatment potently increased Cu toxicity but had no effect on Fe toxicity. These studies demonstrate an important role for neuronal GSH homeostasis in selective protection against Cu toxicity, a finding with widespread implications for neurodegenerative disorders.     

Substrate-bound beta-amyloid peptides inhibit cell adhesion and neurite outgrowth in primary neuronal cultures

Accumulation of the beta-amyloid protein (Abeta) in the brain is an important step in the pathogenesis of Alzheimer's disease. However, the mechanism of Abeta toxicity remains unclear. Abeta can bind to the extracellular matrix, a structure that regulates adhesive events such as neurite outgrowth and synaptogenesis. The binding of Abeta to the extracellular matrix suggests that Abeta may disrupt cell-substrate interactions. Therefore, the effect of substrate-bound Abeta on the growth of isolated chick sympathetic and mouse cortical neurons was examined. Abeta1-40 and Abeta1-42 had dose-dependent effects on cell morphology. When tissue culture plates were coated with 0.1-10 ng/well Abeta, neurite outgrowth increased. Higher amounts of Abeta peptides (> or =3 microg/well) inhibited outgrowth. The inhibitory effect was related to aggregation of the peptide, as preincubation of Abeta1-40 for 24 h at 37 degrees C (a process known to increase amyloid fibril formation) was necessary for inhibition of neurite outgrowth. Abeta29-42, but not Abeta1-28, also inhibited neurite outgrowth at high concentrations, demonstrating that the inhibitory domain is located within the hydrophobic C-terminal region. Abeta1-40, Abeta1-42, and Abeta29-42 also inhibited cell-substrate adhesion, indicating that the effect on neurite outgrowth may have been due to inhibition of cell adhesion. The results suggest that accumulation of Abeta may disrupt cell-adhesion mechanisms in vivo.     

Genistein and daidzein induce neurotoxicity at high concentrations in primary rat neuronal cultures

Summary It is known that estrogen can protect neurons from excitotoxicity. Since isoflavones possess estrogen-like activity, it is of interest to determine whether isoflavones can also protect neurons from glutamate-induced neuronal injury. Morphological observation and lactate dehydrogenase (LDH) release assay were used to estimate the cellular damage. It is surprising that, contrary to estrogen, isoflavones, specifically genistein and daidzein, are toxic to primary neuronal culture at high concentration. Treatment of neurons with 50 μM genistein and daidzein for 24 h increased LDH release by 90% and 67%, respectively, indicating a significant cellular damage. Under the same conditions, estrogen such as 17β-estradiol did not show any effect on primary culture of brain cells. At 100 μM, both genistein and daidzein increased LDH release by 2.6- and 3-fold, respectively with a 30-min incubation. Furthermore, both genistein and daidzein at 50 μM increased the intracellular calcium level, [Ca²⁺]_i, significantly. To determine their mode of action, genistein and daidzein were tested on glutamate and GABA_A receptor binding. Both genistein and daidzein were found to have little effect on glutamate receptor binding, while the binding of [³H]muscimol to GABA_A receptors was markedly inhibited. However, 17β-estradiol did not affect GABA_A receptor binding suggesting that the toxic effect of genistein and daidzein could be due to their inhibition of the GABA_A receptor resulting in further enhancement of excitation by glutamate and leading to cellular damage. Ying Jin, Heng Wu contributed equally to this article.     

Epoxyeicosatrienoic acids enhance axonal growth in primary sensory and cortical neuronal cell cultures

It has recently been reported that soluble epoxide hydrolase (sEH), the major enzyme that metabolizes epoxyeicosatrienoic acids (EETs), is expressed in axons of cortical neurons; however, the functional relevance of axonal sEH localization is unknown. Immunocytochemical analyses demonstrate predominant axonal localization of sEH in primary cultures of not only cortical but also sympathetic and sensory neurons. Morphometric analyses of cultured sensory neurons indicate that exposure to a regioisomeric mixture of EETs (0.01-1.0 μM) causes a concentration-dependent increase in axon outgrowth. This axon promoting activity is not a generalized property of all regioisomers of EETs as axonal growth is enhanced in sensory neurons exposed to 14,15-EET but not 8,9- or 11,12-EET. 14,15-EET also promotes axon outgrowth in cultured cortical neurons. Co-exposure to EETs and either of two structurally diverse pharmacological inhibitors of sEH potentiates the axon-enhancing activity of EETs in sensory and cortical neurons. Mass spectrometry indicates that sEH inhibition significantly increases EETs and significantly decreases dihydroxyeicosatrienoic acid metabolites in neuronal cell cultures. These data indicate that EETs enhance axon outgrowth and suggest that axonal sEH activity regulates EETs-induced axon outgrowth. These findings suggest a novel therapeutic use of sEH inhibitors in promoting nerve regeneration.     

Anandamide-induced cell death in primary neuronal cultures: role of calpain and caspase pathways

Anandamide (arachidonoylethanolamide or AEA) is an endocannabinoid that acts at vanilloid (VR1) as well as at cannabinoid (CB1/CB2) and NMDA receptors. Here, we show that AEA, in a dose-dependent manner, causes cell death in cultured rat cortical neurons and cerebellar granule cells. Inhibition of CB1, CB2, VR1 or NMDA receptors by selective antagonists did not reduce AEA neurotoxicity. Anandamide-induced neuronal cell loss was associated with increased intracellular Ca(2+), nuclear condensation and fragmentation, decreases in mitochondrial membrane potential, translocation of cytochrome c, and upregulation of caspase-3-like activity. However, caspase-3, caspase-8 or caspase-9 inhibitors, or blockade of protein synthesis by cycloheximide did not alter anandamide-related cell death. Moreover, AEA caused cell death in caspase-3-deficient MCF-7 cell line and showed similar cytotoxic effects in caspase-9 dominant-negative, caspase-8 dominant-negative or mock-transfected SH-SY5Y neuroblastoma cells. Anandamide upregulated calpain activity in cortical neurons, as revealed by alpha-spectrin cleavage, which was attenuated by the calpain inhibitor calpastatin. Calpain inhibition significantly limited anandamide-induced neuronal loss and associated cytochrome c release. These data indicate that AEA neurotoxicity appears not to be mediated by CB1, CB2, VR1 or NMDA receptors and suggest that calpain activation, rather than intrinsic or extrinsic caspase pathways, may play a critical role in anandamide-induced cell death.     

Immunocytochemical localization of protein kinase C in developing brain tissue and in primary neuronal cultures

Antisera to protein kinase C (PKC) have been used to examine the presence and distribution of the enzyme in developing cerebellar cortex of postnatal rat and in cultures of rat sympathetic ganglia. In the cerebellar cortex of 2-,4-, and 6-day old rats, immunostaining was observed in areas of early-forming presynaptic terminals and growth cones. No staining was evident in the cortical proliferative zone. Beginning at 10 days postnatal, nuclear staining, not apparent at earlier stages, was prominent in Purkinje cells. In neuronal cultures of dissociated rat sympathetic ganglia, PKC was immunolocalized in cell bodies and bundles of neuronal processes. Immunoreactivity was particularly striking in growth cones of extending neurites and in axonal varicosities. These results suggest a role for PKC in neuronal growth following cell proliferation and in synaptic function. The appearance of nuclear staining in later developmental stages suggests that the enzyme may be involved in the promotion and maintenance of the differentiated state of neurons.     

Prion strain targeting independent of strain-specific neuronal tropism

While neuropathological features that define prion strains include spongiform degeneration and deposition patterns of PrP^Sc, the underlying mechanism for the strain-specific differences in PrP^Sc targeting is not known. To investigate prion strain targeting, we inoculated hamsters in the sciatic nerve with either the hyper (HY) or drowsy (DY) strain of the transmissible mink encephalopathy (TME) agent. Both TME strains were initially retrogradely transported in the central nervous system (CNS) exclusively by four descending motor tracts. The locations of HY and DY PrP^Sc deposition were identical throughout the majority of the incubation period. However, differences in PrP^Sc deposition between these strains were observed upon development of clinical disease. The differences observed were unlikely to be due to strain-specific neuronal tropism, since comparison of PrP^Sc deposition patterns by different routes of infection indicated that all brain areas were susceptible to prion infection by both TME strains. These findings suggest that prion transport and differential susceptibility to prion infection are not solely responsible for prion strain targeting. The data suggest that differences in PrP^Sc distribution between strains during clinical disease are due to differences in the length of time that PrP^Sc has to spread in the CNS before the host succumbs to disease.     

Effects of NMDA and ferrous sulfate on oxidation and cell death in primary neuronal cultures

Excessive oxidative radical production has been implicated in a variety of neurodegerative processes including NMDA (N-methyl-D-aspartate) mediated excitotoxicity. To determine the relationship of oxidation to NMDA-receptor mediated neuronal death, we exposed rat primary cortical neuronal cultures to ferrous sulfate and the fluorescent dyes dichlorofluorescin diacetate (H(2)DCF) and propidium iodide (PI) to monitor reactive oxygen species (ROS) and cell death, respectively in the same cultures. Ferrous sulfate (FeSO(4)) caused a dose-dependent increase in cellular oxidation with an ED(50) of approximately 136 microM. Levels of oxidation increased over time reaching maximum levels between 15 and 25 min. Ferrous sulfate (ED(50) approximately 241 microM) treatment for 25 min caused a delayed and progressive neuronal death that was comparable to NMDA (100 microM, 25 min) delayed neuronal death. NMDA (100 microM, 25 min) alone did not result in measurable increases of DCF fluorescence. However, when combined with 40 microM FeSO(4), NMDA dose-dependently increased H(2)DCF fluorescence. Despite the increase in DCF oxidation, combinations of FeSO(4) with NMDA did not synergize or accelerate NMDA-receptor mediated or glutamate-mediated excitotoxicity. Although excessive amounts FeSO(4) induced oxidation can cause delayed neuronal death, these findings suggest that oxidative stress is not the key factor in triggering the NMDA mediated excitotoxic cascade.     

Precursors of glutamic acid nitrogen in primary neuronal cultures: Studies with15N

We utilized gas chromatography-mass spectrometry to study the transfer of¹⁵N from [2-¹⁵N]glutamine, [¹⁵N]leucine, [¹⁵N]alanine, or¹⁵NH₄Cl to [¹⁵N]glutamate and [¹⁵N]aspartate in cultured cerebrocortical GABA-ergic neurons from the mouse. Initial rates of¹⁵N appearance (atom % excess) were somewhat higher with 2mM [2-¹⁵N]glutamine as a precursor than with 1mM [¹⁵N]leucine or 1mM [¹⁵N]alanine, but initial net formation (nmol [¹⁵N]glutamate/mg protein.min^–1) was roughly comparable with all precursors. At steady-state¹⁵N labeling was about two times greater with 2mM [2-¹⁵N]glutamine as precursor. The subsequent transfer of¹⁵N from glutamate to aspartate was extremely rapid, the labelling pattern of these two amino acid pools being virtually indistinguishable. We observed little reductive amination of 2-oxo-glutarate to yield [¹⁵N]glutamate in the presence of 0.3mM¹⁵NH₄Cl. Reductive amination through glutamate dehydrogenase was much more prominent at a concentration of 3.0mM¹⁵NH₄Cl. Glutamate formation via reductive amination was unaffected by inclusion of 1 mM 2-oxo-glutarate in the incubation medium. These results indicate that glutamate synthesis in cultured GABA-ergic neurons is derived not only from the glutaminase reaction, but also from transamination reactions in which both leucine and alamine are efficient N donors. Reductive amination of 2-oxo-glutarate in the glutamate dehydrogenase pathway plays a relatively minor role at lower concentrations of extracellular ammonia but becomes quite active at 3mM ammonia.     

Immunocytochemical localization of glycogen phosphorylase kinase in rat brain sections and in glial and neuronal primary cultures

Summary. The physiological function of brain glycogen and the role of phosphorylase kinase as a regulatory enzyme in the cascade of reactions associated with glycogenolysis in the brain have not been fully elucidated. As a first step toward elucidating such a function, we studied the localization of phosphorylase kinase in glial and neuronal primary cell cultures, and in adult rat brain slices, using a rabbit polyclonal antibody against skeletal muscle glycogen phosphorylase kinase. Immunocytochemical examination of rat astroglia-rich primary cultures revealed that a large number of cells were positive for glycogen phosphorylase kinase immunoreactivity. These cells were also positive for vimentin, a marker for immature glia, while they were negative for glial fibrillary acidic protein, a marker for mature astroglia, and for galactocerebroside, an oligodendroglial marker. Neurons in rat neuron-rich primary cultures did not show any kinase-positive staining. In paraformaldehyde-fixed adult rat brain sections, phosphorylase kinase immunoreactivity was detected in glial-like cells throughout the brain, with relatively high staining found in the cerebral cortex, the cerebellum, and the medulla oblongata. Phosphorylase kinase immunoreactivity could not be detected in neurons, with the exception of a group of large neurons in the brain stem, most likely belonging to the mesencephalic trigeminal nucleus. Phosphorylase kinase was also localized in the choroid plexus and to a lesser degree in the ependymal cells lining the ventricles. Phosphorylase kinase thus appears to have the same cellular distribution in nervous tissue as its substrates, i.e. glycogen phosphorylase and glycogen, which suggests that the physiological role of brain phosphorylase kinase is the mobilization of glycogen stores to fuel the increased metabolic demands of neurons and astrocytes.