The effect of microglia on embryonic dopaminergic neuronal survival in vitro: diffusible signals from neurons and glia change microglia from neurotoxic to neuroprotective

Trisialoganglioside GT1b induces in vivo degeneration of nigral dopaminergic neurons: role of microglia

We recently showed that trisialoganglioside (GT1b) induces cell death of dopaminergic neurons in rat mesencephalic cultures (Chung et al., Neuroreport 12:611-614, 2001). The present study examines the in vivo neurotoxic effects of GT1b on dopaminergic neurons in the substantia nigra (SN) of Sprague-Dawley rats. Seven days after GT1b injection into the SN, immunocytochemical staining of SN tissue revealed death of nigral neurons, including dopaminergic neurons. Additional immunostaining using OX-42 and OX-6 antibodies showed that GT1b-activated microglia were present in the SN where degeneration of nigral neurons was found. Western blot analysis and double-labeled immunohistochemistry showed that inducible nitric oxide synthase (iNOS) was expressed in the SN, where its levels were maximal at 8 h post-GT1b injection, and that iNOS was localized exclusively within microglia. GT1b-induced loss of dopaminergic neurons in the SN was partially inhibited by N(G)-nitro-L-arginine methyl ester hydrochloride, an NOS inhibitor. Our results indicate that in vivo neurotoxicity of GT1b against nigral dopaminergic neurons is at least in part mediated by nitric oxide released from activated microglia. Because GT1b exists abundantly in central nervous system neuronal membranes, our data support the hypothesis that immune-mediated events triggered by endogenous compounds such as GT1b could contribute to the initiation and/or the progression of dopaminergic neuronal cell death that occurs in Parkinson's disease.

Phenotypic and functional changes in glial cells as a function of age

Both in vivo and in vitro investigations point to an important role for the immune system in the development of age-related neurodegeneration. Microglia isolated from aged female F344 rats, 18-20 months, show a higher percentage of cells with an ameboid morphology indicative of activation, whereas, astrocytes had a quiescent morphology. The ability of astrocytes and microglia to attenuate toxin-induced neuronal injury was examined. Post-natal day 1-3 pup cells optimally rescued neurons from Abeta-induced toxicity, whereas mixed glial cells from 18-20 month old rats were unable to rescue neurons from Abeta-induced toxicity. Our results suggested the appearance of a neurotoxic co-factor, therefore we investigated the basal level of nitric oxide and pro-inflammatory cytokines to determine if altered levels of immune mediators play a role in the toxicity. Mitogen-stimulated nitric oxide production increased 10 fold with age of donor, whereas, only the pup cells expressed an increase in TNF-alpha production. Basal levels of pro-inflammatory cytokines, as measured by RNA protection assays, increased with age. In particular, IL-1beta was increased 2 fold between adult and aged glial cells. The elevated cytokine expression may contribute to enhanced susceptibility to neurodegenerative diseases.

Microglial reaction induced by noncytotoxic methylmercury treatment leads to neuroprotection via interactions with astrocytes and IL-6 release

Microglial cells react early to a neurotoxic insult. However, the bioactive factors and the cell-cell interactions leading to microglial activation and finally to a neuroprotective or neurodegenerative outcome remain to be elucidated. Therefore, we analyzed the microglial reaction induced by methylmercury (MeHgCl) using cell cultures of different complexity. Isolated microglia were found to be directly activated by MeHgCl (10(-10) to 10(-6) M), as indicated by process retraction, enhanced lectin staining, and cluster formation. An association of MeHgCl-induced microglial clusters with astrocytes and neurons was observed in three-dimensional cultures. Close proximity was found between the clusters of lectin-stained microglia and astrocytes immunostained for glial fibrillary acidic protein (GFAP), which may facilitate interactions between astrocytes and reactive microglia. In contrast, immunoreactivity for microtubule-associated protein (MAP-2), a neuronal marker, was absent in the vicinity of the microglial clusters. Interactions between astrocytes and microglia were studied in cocultures treated for 10 days with MeHgCl. Interleukin-6 release was increased at 10(-7) M of MeHgCl, whereas it was decreased when each of these two cell types was cultured separately. Moreover, addition of IL-6 to three-dimensional brain cell cultures treated with 3 x 10(-7) M of MeHgCl prevented the decrease in immunostaining of the neuronal markers MAP-2 and neurofilament-M. IL-6 administered to three-dimensional cultures in the absence of MeHgCl caused astrogliosis, as indicated by increased GFAP immunoreactivity. Altogether, these results show that microglial cells are directly activated by MeHgCl and that the interaction between activated microglia and astrocytes can increase local IL-6 release, which may cause astrocyte reactivity and neuroprotection.

Microglial cells protect cerebellar granule neurons from apoptosis: evidence for reciprocal signaling

The microglia are the immune cell population of the nervous system and play important roles both in normal function and in disease. Reciprocal neuron-microglia interactions are not well understood, in particular those concerning the crosstalk between the two cell populations when neuronal damage does occur. We have used a well-established model of apoptosis in cerebellar granule neurons to test the effect of co-culturing microglial cells with them or of exposing them to microglia-conditioned medium. Microglial cells, derived from cortical or cerebellar mixed glial cultures and plated over cerebellar granule neurons, protected these neurons from apoptosis induced by shifting them, at 7 days in vitro, for 24 h from a depolarizing (high-potassium) to a nondepolarizing (low-potassium) medium. The same result was achieved when microglial cells obtained from mixed glial cortical cultures were plated over a membrane well insert in the culture chamber, permitting medium exchange without physical contact with granule neurons. A similar result was obtained when the low-potassium, apoptosis-inducing medium was conditioned by 48-h exposure to microglial cells; 24-h exposure to microglial cells was not enough to confer neuroprotective capability to the conditioned medium. However in double-conditioned medium experiments, in which the medium was first exposed to apoptotic neurons and then to microglial cells, unknown signal(s) released by apoptotic neurons, conferred to the 24-h conditioned medium a strong neuroprotective action, similar to that observed in the co-cultures experiments. This finding, together with the results from co-culture experiments, is explained by admitting that molecules released in the medium by apoptotic neurons potentiate the anti-apoptotic activity of microglia. Our results, therefore, demonstrate not only that normally microglial cells release in the medium molecule(s) able to rescue neurons from apoptotic death, but that unknown diffusible signal(s) from apoptotic neurons enhance(s) microglial neuroprotective properties as well.

Inhibition of tumour necrosis factor-alpha (TNFalpha)-induced NF-kappaB p52 converts the metabolic effects of microglial-derived TNFalpha on mouse cerebellar neurones to neurotoxicity

Activated microglia are implicated in the injury of neurones and macroglia both in vitro and in vivo. Here, we demonstrate that media conditioned by interferon-gamma treated microglia initially impair the metabolism of mouse cerebellar neurones grown in serum-free conditions without inducing cell death. Metabolic effects include inhibition of the ability of mitochondria to reduce 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) and cytochrome oxidase activity. These effects are blocked by antibodies to tumour necrosis factor-alpha (TNFalpha), a cytokine produced by microglial activation, and they are not reproduced by media conditioned by resting microglia. The metabolic effects are evident for up to 24 h in vitro. More prolonged exposure, up to 48 h, results in TNFalpha dependent neuronal death as previously observed. Between 2 and 48 h TNFalpha present in media conditioned by interferon-gamma treated but not resting microglia is associated with nuclear factor kappa B (NF-kappaB) consensus sequence binding in paired mouse cerebellar neuronal cultures without affecting activation of the signal transducer and activator of transcription (STAT) transcription factor. Neuronal death can be accelerated by peptide blockade of the nuclear transport of NF-kappaB p52 subunit during exposure of cerebellar neurones to medium from interferon-gamma treated microglia. This toxicity is blocked by anti-TNFalpha antibody. Soluble factors released by activated microglia therefore contribute to neuronal dysfunction that is initially reversible but may culminate in neurotoxicity. Characterizing and manipulating these events in vivo theoretically provides an opportunity for neuroprotection in selected diseases affecting the central nervous system.

Naloxone prevents microglia-induced degeneration of dopaminergic substantia nigra neurons in adult rats

Resident microglia are involved in immune responses of the central nervous system and may contribute to neuronal degeneration and death. Here, we tested in adult rats whether injection of bacterial lipopolysaccharide (which causes inflammation and microglial activation) just above the substantia nigra, results in the death of dopaminergic substantia nigra pars compacta neurons. Two weeks after lipopolysaccharide injection, microglial activation was evident throughout the nigra and the number of retrogradely-labeled substantia nigra neurons was reduced to 66% of normal. This suggests that inflammation and/or microglial activation can lead to neuronal cell death in a well-defined adult animal model. The opioid receptor antagonist naloxone reportedly reduces release of cytotoxic substances from microglia and protects cortical neurons in vitro. Here, a continuous two-week infusion of naloxone at a micromolar concentration close to the substantia nigra, prevented most of the neuronal death caused by lipopolysaccharide, i.e. 85% of the neurons survived. In addition, with systemic (subcutaneous) infusion of 0. 1mg/d naloxone, 94% of the neurons survived. Naloxone infusions did not obviously affect the morphological signs of microglial activation, suggesting that naloxone reduces the release of microglial-derived cytotoxic substances. Alternatively, microglia might not cause the neuronal loss, or naloxone might act by blocking opioid receptors on (dopaminergic or GABAergic) neurons.Thus, local inflammation induces and the opioid antagonist naloxone prevents the death of dopaminergic substantia nigra neurons in adult rats. This may be relevant to the understanding of the pathology and treatment of Parkinson's disease, where these neurons degenerate.

Additive effects of caspase inhibitor and lazaroid on the survival of transplanted rat and human embryonic dopamine neurons

Major practical constraints on neural grafting in Parkinson's disease are the shortage of human donor tissue and the great loss of dopamine neurons during the grafting procedure. The vast majority of implanted embryonic dopamine neurons are believed to die within a few days of transplantation surgery, at least in part through apoptosis. We have previously found that survival of nigral grafts in rodents can be significantly augmented by pretreatment with the caspase inhibitor Ac-YVAD-cmk or by lazaroids (lipid peroxidation inhibitors). We now report that pretreatment with the caspase inhibitor Ac-DEVD-cmk, but not z-VAD-fmk, results in a significantly improved survival of transplanted dopamine neurons of similar magnitude to that achieved in this study using Ac-YVAD-cmk (both 220-230% of control). In addition, we found that treatment of the graft tissue with tirilazad mesylate (a lazaroid allowed for clinical use) almost doubled the survival of grafted dopamine neurons. When Ac-YVAD-cmk and tirilazad mesylate treatments were combined, the number of surviving dopamine neurons increased significantly further to 280% of control. Importantly, the same combination of neuroprotectants enhanced the survival of human dopamine neurons xenotransplanted to immunosuppressed rats (to 240% of control). In conclusion, these results suggest that combining treatments that counteract oxidative stress and caspase activation is a valuable strategy to enhance nigral graft survival that should be considered for clinical application.

Microglia in organotypic hippocampal slice culture and effects of hypoxia: ultrastructure and lipocortin-1 immunoreactivity

Lipocortin-1 immunocytochemistry was used to study the various cell forms of microglia that appear during organotypic hippocampal tissue culture, as well as in the in vitro toxic hypoxia model. Antibodies against lipocortin-1 identified activated and phagocytic cells that were abundant in a slice after the plating of a culture: cells of the intermediate form at the later time-points of culturing, resting ramified microglia beginning from the seventh day of culturing, as well as activated and phagocytic cells that appeared in the slice after experimental toxic hypoxia induced by potassium cyanide treatment. Lipocortin-1-positive microglia cell forms corresponded well to the description of the microglia in vivo, and the morphology of microglia corresponded to the circumstances under which these cells were observed in slice cultures. Electron microscopic studies have demonstrated, for the first time, that microglia in organotypic slice culture preserve morphological features typical of different microglial forms in vivo, as well as specific contacts and interactions with the other neural tissue elements. After experimental toxic hypoxia, rapid changes in microglial ultrastructure and localization were observed, reminiscent of in vivo models of ischaemia. In conclusion, observations of microglial morphology and behaviour allow us to suggest that microglia in the organotypic culture preserve their essential characteristic features and properties, thus providing an important model system for studying the structure and function of these cells.

Impact of a preceding striatal excitotoxic lesion and treatment with ciliary neurotrophic factor on striatal graft survival

The survival of grafted embryonic striatal tissue, dissected from the lateral ganglionic eminence, depends on the status of the host striatum. We found significantly larger volumes of surviving graft tissue and of striatal-like tissue (P-zone) within the graft, when the host striatum had been subjected to an excitotoxic lesion prior to transplantation surgery. Concomitantly the numbers of surviving grafted cells, assessed in both cresyl violet-stained sections and in sections stained with an immunohistochemical marker for striatal neurons, increased as compared to when graft tissue was placed in an intact unlesioned striatum. Finally, we examined the impact of treatment of the donor tissue with ciliary neurotrophic factor (CNTF) on graft survival. CNTF has previously been shown to protect striatal neurons against excitotoxic insults both in vitro and in vivo, but it did not improve striatal graft survival when added to the cell suspension prior to implantation.

The glial scar and central nervous system repair

Damage to the central nervous system (CNS) results in a glial reaction, leading eventually to the formation of a glial scar. In this environment, axon regeneration fails, and remyelination may also be unsuccessful. The glial reaction to injury recruits microglia, oligodendrocyte precursors, meningeal cells, astrocytes and stem cells. Damaged CNS also contains oligodendrocytes and myelin debris. Most of these cell types produce molecules that have been shown to be inhibitory to axon regeneration. Oligodendrocytes produce NI250, myelin-associated glycoprotein (MAG), and tenascin-R, oligodendrocyte precursors produce NG2 DSD-1/phosphacan and versican, astrocytes produce tenascin, brevican, and neurocan, and can be stimulated to produce NG2, meningeal cells produce NG2 and other proteoglycans, and activated microglia produce free radicals, nitric oxide, and arachidonic acid derivatives. Many of these molecules must participate in rendering the damaged CNS inhibitory for axon regeneration. Demyelinated plaques in multiple sclerosis consists mostly of scar-type astrocytes and naked axons. The extent to which the astrocytosis is responsible for blocking remyelination is not established, but astrocytes inhibit the migration of both oligodendrocyte precursors and Schwann cells which must restrict their access to demyelinated axons.