Enhancer II-targeted dsRNA decreases GDNF expression via histone H3K9 trimethylation to inhibit glioblastoma progression.

BACKGROUND: Glial cell line-derived neurotrophic factor (GDNF) is expressed in both astrocytes and glioblastoma (GBM) cells. GDNF expression is significantly increased in GBM, and inhibiting its expression can retard GBM progression. However, there is no known method for specific inhibition of GDNF in GBM cells. METHODS: Promoter-targeted dsRNA-induced transcriptional gene silencing or activation was recently achieved in human cells. This approach has the potential to specifically regulate gene transcription via epigenetic modifications. In this study, we designed six candidate dsRNAs targeting the enhancer or silencer in GDNF gene promoter II to check their effects on GDNF transcription and GBM progression. RESULTS: Among these dsRNAs, enhancer II-targeted dsRNA significantly inhibited U251 GBM progression by downregulating GDNF (P < 0.05), while silencer II-targeted dsRNA exerted an opposite effect. Moreover, enhancer II-targeted dsRNA did not significantly change GDNF expression in human astrocytes (HA) and the proliferation and migration of HA cells (P > 0.05). Bisulfate PCR and chromatin immunoprecipitation analyses revealed that both DNA methylation and trimethylation of histone 3 at lysine 9 (H3K9me3) at silencer II-targeted region significantly increased, and H3K9me3 at enhancer II-targeted region significantly decreased, in U251 cells compared with HA cells in non-intervention condition (P < 0.05). Both enhancer II- and silencer II-targeted dsRNA significantly increased H3K9me3 methylation rather than DNA at the targeted site in U251 cells (P < 0.05). The expression and activity of histone methyltransferase SETDB1 increased dramatically in U251 cells compared with HA cells, and it was recruited to enhancer II targeting region after enhancer II-targeted dsRNA treatment in U251 cells (P < 0.05). CONCLUSIONS: Our results demonstrate that a promoter-targeted dsRNA can silence or promote gene transcription depending on its targeted site in different cis-acting elements in the gene promoter. Targeted inhibition of GDNF by enhancer II-targeted dsRNA may be explored as a novel treatment for GBM.

Down-regulation of microRNA-216a confers protection against yttrium aluminium garnet laser-induced retinal injury via the GDNF-mediated GDNF/GFRα1/RET signalling pathway.

Retinal injury plays a leading role in the onset of visual impairment. Current forms of treatment are not able to ameliorate scarring, cell death and tissue and axon regeneration. Recently, microRNA-216a (miR-216a) has been reported to regulate snx5, a novel notch signalling pathway component during retinal development. This study aims to elucidate the role of miR-216a in yttrium aluminium garnet (YAG) laser-induced retinal injury by targeting glial cell line-derived neurotrophic factor (GDNF) via GDNF/GDNF family neurotrophic factor receptor α1 (GFRα1)/rearranged during transfection (RET) signalling pathway. Wistar male rats were first randomly assigned into control and model groups. Immunohistochemistry was performed to detect the GDNF positive expression rate and terminal deoxynucleotidyl transferase-mediated dUTP nick end labelling (TUNEL) staining for apoptotic index (AI) of retinal tissue. Retinal neurons were divided into normal, blank, negative control (NC), miR-216a mimic, miR-216a inhibitor, siRNA-GDNF and miR-216a inhibitor?siRNA-GDNF groups. Dual luciferase reporter assay was conducted in order to identify the targeting relationship between GDNF and miR-216a. Reverse transcription quantitative polymerase chain reaction (RT-qPCR) and western blot were used for the analysis of mRNA and protein levels of miR-216a and related genes. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was used to determine cell proliferation and flow cytometry was used to observe cell cycle and apoptosis. Results show that the model group had an increased GDNF positive rate, AI of retinal tissue and mRNA and protein levels of cellular oncogene fos (c-fos), vascular endothelial growth factor (VEGF), brain-derived neurotrophic factor (BDNF), GDNF, GFRα1 and bcl-2-associated X protein (bax), declined miR-216a level and mRNA and protein levels of RET and bcl-2 compared with the control group. GDNF was verified as the target gene for miR-216a. Compared with the blank and NC groups, the miR-216a mimic and siRNA-GDNF groups had higher mRNA and protein levels of c-fos, VEGF and bax, cell number in the G1 phase and increased cell apoptosis but reduced BDNF, GDNF, GFRα1, RET and bcl-2 expression, cell proliferation and cell numbers in the S phase, while the opposite trend was observed in the miR-216a inhibitor group. Taken together, our findings demonstrate that elevated GDNF levels can reduce the retinal injury, whereby down-regulated miR-216a aggravates the YAG laser-induced retinal injury by targeting the GDNF level through the GDNF/ GFRα1/RET signalling pathway.

Targeting the divergent TGFß superfamily cytokine MIC-1/GDF15 for therapy of anorexia/cachexia syndromes.

PURPOSE OF REVIEW: To review recent finding on MIC-1/GDF15 and re-evaluate it as a potential target for the therapy of anorexia/cachexia syndromes. RECENT FINDINGS: MIC-1/GDF15 consistently induces anorexia/cachexia in animal models. Its actions on brainstem feeding centers leads to anorexia, inducing prolonged undernutrition and consequent loss of both lean and fat mass. Epidemiological studies by multiple groups have linked substantially elevated serum levels of this cytokine to anorexia/cachexia syndromes in diverse diseases such as cancer, chronic renal and cardiac failure, and chronic obstructive lung disease. These elevated serum levels are similar to those required to induce this syndrome in animals. Recent identifications of its previously elusive receptor as GFRAL, has enhanced understanding of its biology and suggests that modulating the MIC-1/GDF15-GFRAL pathway may be a therapeutic target for anorexia/cachexia syndrome. SUMMARY: Inhibiting MIC-1/GDF15 or its receptor GFRAL are high-value potential targets for treatment of anorexia/cachexia syndrome in patients whose elevated serum levels may justify its use.

The effect of amino plasma-enhanced chemical vapor deposition-treated titanium surface on Schwann cells.

The surface modification of titanium and its alloys with amino group plasma-enhanced chemical vapor deposition has been proven to enhance the performance of implants on initial osteoblast bioactivity in vitro. However, scarce information on the effect of this kind of surface modification on nerve regeneration exists. In this study, the surface chemistry of pure Ti disks and surface-modified disks was examined using X-ray photoelectron spectroscopy. Cell counting kit 8 assay, 4,6-diamidino-2-phenylindole staining, flow cytometry, and scanning electron microscopy showed that either the p30% or cw + p30% mode-mediated surface significantly promote Schwann cell adhesion without any cytotoxicity compared with the pure Ti surface, and the cw + p30% group showed the best performance on cell adhesion. However, results of polymerase chain reaction and Western blot analyses showed that the mRNA and protein levels of glial cell-derived neurotrophic factor and nerve growth factor of the p30% and cw + p30% groups were lower than those of the Ti group at some time points. Generally, the results indicate that amino-functionalized Ti surfaces can promote Schwann cell adhesion without cytotoxicity, but this modification, in fact, inhibited the expression of the key growth factors GDNF and NGF of Schwann cells. © 2017 Wiley Periodicals, Inc. J Biomed Mater Res Part A: 106A: 265-271, 2018.

Docosahexaenoic acid induces glial cell-line derived neurotrophic factor release in C6 glioma cells: Implications of antidepressant effects for docosahexaenoic acid.

Dietary deficiency of n-3 polyunsaturated fatty acids (PUFAs) is involved in the pathophysiology and etiology of major depressive disorder. Supplementation with docosahexaenoic acid (DHA) exerts antidepressant-like effect; however, the molecular mechanism of DHA action remains unclear. Here we examined the effects of DHA on the modulation of glial cell line-derived neurotrophic factor (GDNF), which is essential for neural development, plasticity, neurogenesis, and survival. We demonstrated that DHA treatment significantly increased GDNF release in a concentration dependent manner in rat C6 glioma cells (C6 cells) and primary cultured rat astrocytes, which is also associated with increased expression of GDNF mRNA. Furthermore, the DHA-induced GDNF production was inhibited by mitogen activated protein kinase (MEK) inhibitor and protein kinase C (PKC) inhibitor, but not protein kinase A (PKA) inhibitor and p38 mitogen-activated protein kinase (MAPK) inhibitor. DHA-induced extracellular signal-regulated kinase (ERK) activation is dependent on the PKC, as demonstrated by its reversibility after pretreatment with PKC inhibitor. Moreover, fibroblast growth factor receptor (FGFR inhibitor) but not epidermal growth factor receptor (EGFR) inhibitor blocked the activation of ERK induced by DHA treatment. DHA-induced GDNF production was also blocked by FGFR inhibitor, suggesting that FGFR is also involved in ERK activation-mediated GDNF production induced by DHA. Our study demonstrates that DHA-induced release of GDNF, mediated by PKC and FGFR-dependent on ERK activation, may contribute to the antidepressant-like effect of DHA.

Blockade of glial-derived neurotrophic factor in laryngeal muscles promotes appropriate reinnervation.

OBJECTIVES/HYPOTHESIS: Synkinetic reinnervation of the laryngeal muscles is one of the causes of the poor functional recovery after a recurrent laryngeal nerve (RLN) injury. Glial-derived neurotrophic factor (GDNF) is elevated in rat laryngeal muscles during RLN reinnervation. The specific aim of this investigation was to evaluate the effect of anti-GDNF on RLN reinnervation. METHODS: Anti-GDNF antibody was injected into the posterior cricoarytenoid (PCA) 3 days following RLN transection and anastomosis. Larynges were harvested at 7, 14, 28, 56, and 112 days post injury (DPI). Prior to sacrifice, the vocal fold mobility was assessed. Immunostaining to identify neuromuscular junctions was used to evaluate the extent of axonal reinnervation of the PCA, lateral thyroarytenoid (LTA), and medial thyroarytenoid (MTA). RESULTS: After anti-GDNF injection into PCA, RLN reinnervation in all muscles was altered when compared to the controls. PCA innervation was delayed. At 7 DPI, only a few axons made synapses in the PCA. In contrast, axons prematurely innervated the LTA and MTA when compared to controls. Innervation was similar to controls at 56 and 112 DPI. Vocal fold motion was enhanced in 10 of 24 animals studied. CONCLUSIONS: After injection of anti-GDNF into the PCA, early arriving axons bypass the PCA and enter the LTA. Later arriving axons innervate the PCA and MTA. Vocal fold function is improved as compared to controls. Anti-GDNF injection into the PCA influences the pattern of reinnervation and may result in less synkinetic, more functional innervation. LEVEL OF EVIDENCE: NA Laryngoscope, 126:E337-E342, 2016.

Sonic Hedgehog Controls the Phenotypic Fate and Therapeutic Efficacy of Grafted Neural Precursor Cells in a Model of Nigrostriatal Neurodegeneration.

The expression of soluble growth and survival promoting factors by neural precursor cells (NPCs) is suggested to be a prominent mechanism underlying the protective and regenerative effects of these cells after transplantation. Nevertheless, how and to what extent specific NPC-expressed factors contribute to therapeutic effects is not well understood. Using RNA silencing, the current study investigated the roles of two donor NPC molecules, namely glial cell-line derived neurotrophic factor (GDNF) and sonic hedgehog (SHH), in the protection of substantia nigra dopamine neurons in rats treated with 6-hydroxydopamine (6-OHDA). Analyses indicate that as opposed to the knock-down of GDNF, SHH inhibition caused a profound decline in nigrostriatal neuroprotection. Further, SHH silencing also curbed endogenous neurogenesis and the migration of host brdU+/dcx+ neural precursors into the striatum, which was present in the animals receiving control or GDNF silenced NPCs. A change in graft phenotype, mainly reflected by a reduced proportion of undifferentiated nestin+ cells, as well as a significantly greater host microglial activity, suggested an important role for these processes in the attenuation of neuroprotection and neurogenesis upon SHH silencing. Overall these studies reveal core mechanisms fundamental to grafted NPC-based therapeutic effects, and delineate the particular contributions of two graft-expressed molecules, SHH and GDNF, in mediating midbrain dopamine neuron protection, and host plasticity after NPC transplantation.

Intestinal smooth muscle phenotype determines enteric neuronal survival via GDNF expression.

Intestinal inflammation causes initial axonal degeneration and neuronal death, as well as the proliferation of intestinal smooth muscle cells (ISMC), but subsequent axonal outgrowth leads to re-innervation. We recently showed that expression of glial cell-derived neurotrophic factor (GDNF), the critical neurotrophin for the post-natal enteric nervous system (ENS) is upregulated in ISMC by inflammatory cytokines, leading us to explore the relationship between ISMC growth and GDNF expression. In co-cultures of myenteric neurons and ISMC, GDNF or fetal calf serum (FCS) was equally effective in supporting neuronal survival, with neurons forming extensive axonal networks among the ISMC. However, only GDNF was effective in low-density cultures where neurons lacked contact with ISMC. In early-passage cultures of colonic circular smooth muscle cells (CSMC), polymerase chain reaction (PCR) and western blotting showed that proliferation was associated with expression of GDNF, and the successful survival of neonatal neurons co-cultured on CSMC was blocked by vandetanib or siGDNF. In tri-nitrobenzene sulfonic acid (TNBS)-induced colitis, immunocytochemistry showed the selective expression of GDNF in proliferating CSMC, suggesting that smooth muscle proliferation supports the ENS in vivo as well as in vitro. However, high-passage CSMC expressed significantly less GDNF and failed to support neuronal survival, while expressing reduced amounts of smooth muscle marker proteins. We conclude that in the inflamed intestine, smooth muscle proliferation supports the ENS, and thus its own re-innervation, by expression of GDNF. In chronic inflammation, a compromised smooth muscle phenotype may lead to progressive neural damage. Intestinal stricture formation in human disease, such as inflammatory bowel disease (IBD), may be an endpoint of failure of this homeostatic mechanism.

Mildronate as a regulator of protein expression in a rat model of Parkinson's disease.

BACKGROUND: Mildronate (3-[2,2,2-trimethylhydrazinium] propionate dihydrate) traditionally is a well-known cardioprotective drug. However, our recent studies convincingly demonstrated its neuroprotective properties. The aim of the present study was to evaluate the influence of mildronate on the expression of proteins that are involved in the differentiation and survival of the nigrostriatal dopaminergic neurons in the rat model of Parkinson's disease (PD). The following biomarkers were used: heat shock protein 70 (Hsp70, a molecular chaperone), glial cell line-derived nerve growth factor (GDNF, a growth factor promoting neuronal differentiation, regeneration, and survival), and neural cell adhesion molecule (NCAM). MATERIAL AND METHODS: PD was modeled by 6-hydroxydopamine (6-OHDA) unilateral intrastriatal injection in rats. Mildronate was administered at doses of 10, 20, and 50 mg/kg for 2 weeks intraperitoneally before 6-OHDA injection. Rat brains were dissected on day 28 after discontinuation of mildronate injections. The expression of biomarkers was assessed immunohistochemically and by western blot assay. RESULTS: 6-OHDA decreased the expression of Hsp70 and GDNF in the lesioned striatum and substantia nigra, whereas in mildronate-pretreated (20 and 50 mg/kg) rats, the expression of Hsp70 and GDNF was close to the control group values. NCAM expression also was decreased by 6-OHDA in the striatum and it was totally protected by mildronate at a dose of 50 mg/kg. In contrast, in the substantia nigra, 6-OHDA increased the expression of NCAM, while mildronate pretreatment (20 and 50 mg/kg) reversed the 6-OHDA-induced overexpression of NCAM close to the control values. CONCLUSION: The obtained data showed that mildronate was capable to regulate the expression of proteins that play a role in the homeostasis of neuro-glial processes.

Growth factors and neuropathic pain.

A treatment for neuropathic pain is an important unmet medical need because this pain often is refractory to many medical interventions. An important element in the development of neuropathic pain is a dysfunction in the activity of peripheral nerves. Because neurotrophic factors affect nerve development and maintenance, modulating the activity of these factors can alter neuronal pathophysiology and produce a disease-modifying effect. Blocking the activity of nerve growth factor or enhancing the activity of either glial-derived neurotrophic factor or artemin has shown potential for normalizing neuronal activity and attenuating signs of neuropathic pain in animal models and clinical studies. This article discusses the role of these factors in neuropathic pain and the implications for the development of novel therapeutics.

Role of ventral tegmental area glial cell line-derived neurotrophic factor in incubation of cocaine craving.

BACKGROUND: Ventral tegmental area (VTA) brain-derived neurotrophic factor (BDNF) contributes to time-dependent increases in cue-induced cocaine seeking after withdrawal (incubation of cocaine craving). Here, we studied the role of glial cell line-derived neurotrophic factor (GDNF) in incubation of cocaine craving because, like BDNF, GDNF provides trophic support to midbrain dopamine neurons. METHODS: We first trained rats to self-administer intravenous cocaine for 10 days (6 hours/d, cocaine injections were paired with a tone-light cue). We then manipulated VTA GDNF function and assessed cue-induced cocaine seeking in extinction tests after withdrawal from cocaine. RESULTS: VTA injections of an adeno-associated virus (AAV) vector containing rat GDNF cDNA (5 x 10(8) viral genomes) on withdrawal Day 1 increased cue-induced cocaine seeking on withdrawal days 11 and 31; this effect was not observed after VTA injections of an AAV viral vector containing red fluorescent protein (RFP). Additionally, VTA, but not substantial nigra (SN), GDNF injections (1.25 microg or 12.5 microg/side) immediately after the last cocaine self-administration session increased cue-induced drug seeking on withdrawal days 3 and 10; this effect was reversed by VTA injections of U0126, which inhibits the activity of extracellular signal-regulated kinases (ERK). Finally, interfering with VTA GDNF function by chronic delivery of anti-GDNF monoclonal neutralizing antibodies via minipumps (600 ng/side/d) during withdrawal Days 1-14 prevented the time-dependent increases in cue-induced cocaine seeking on withdrawal days 11 and 31. CONCLUSIONS: Our results indicate that during the first weeks of withdrawal from cocaine self-administration, GDNF-dependent neuroadaptations in midbrain VTA neurons play an important role in the development of incubation of cocaine craving.

Downregulation of Spry-1, an inhibitor of GDNF/Ret, causes angiotensin II-induced ureteric bud branching.

Mutations of genes in the renin-angiotensin system are associated with congenital abnormalities of the kidney and urinary tract. The major signaling pathway for branching morphogenesis during kidney development is the c-Ret receptor tyrosine kinase whose ligand is GDNF and whose downstream target is Wnt11. We determined whether angiotensin II, an inducer of ureteric bud branching in vitro, influences the GDNF/c-Ret/Wnt11 pathway. Mouse metanephroi were grown in the presence or absence of angiotensin II or an angiotensin type 1 receptor (AT1R) antagonist and gene expression was measured by whole mount in situ hybridization. Angiotensin II induced the expression of c-Ret and Wnt11 in ureteric bud tip cells. GDNF, a Wnt11-regulated gene expressed in the mesenchyme, was also upregulated by angiotensin II but this downregulated Spry1, an endogenous inhibitor of Ret tyrosine kinase activity in an AT1R-dependent manner. Angiotensin II also decreased Spry1 mRNA levels in cultured ureteric bud cells. Exogenous angiotensin II preferentially stimulated ureteric bud tip cell proliferation in vivo while AT1R blockade increased cell apoptosis. Our findings suggest AT1R-mediated inhibition of the Spry1 gene increases c-Ret tyrosine kinase activity leading to upregulation of its downstream target Wnt11. Enhanced Wnt11 expression induces GDNF in adjacent mesenchyme causing focal bursts of ureteric bud tip cell proliferation, decreased tip cell apoptosis and branching.

Activation of microglia by endotoxin suppresses the secretion of glial cell line-derived neurotrophic factor (GDNF) through the action of protein kinase C alpha (PKCalpha) and mitogen-activated protein kinases (MAPKS).

The ability of microglia to produce/secrete glial cell line-derived neurotrophic factor (GDNF) in vitro was examined. Immunoblotting analysis revealed that nonstimulated microglia release limited amounts of GDNF with molecular sizes of 14 and 17 kDa. However, the secreted amounts significantly decreased when the microglia were activated with the endotoxin lipopolysaccharide (LPS). Comparison of the amounts of GDNF in the cells and the conditioned medium between the nonstimulated microglia and LPS-stimulated microglia clarified that the secretion of GDNF, but not its production, is strongly suppressed when the microglia are activated with LPS. The inhibitor experiments suggested that the GDNF secretion is depressed by a signaling cascade associated with protein kinase C alpha (PKCalpha) and/or mitogen-activated protein kinases (MAPKs). As expected from the above results, a PKC activator suppressed the secretion of GDNF in nonstimulated microglia. Taken together, these results demonstrated that microglia have the ability to produce and secrete GDNF in vitro, and that the secretion is suppressed by stimulation with endotoxin, probably due to a signaling mechanism involving PKCalpha and/or MAPKs.

Interleukin-1beta mediates GDNF up-regulation upon dopaminergic injury in ventral midbrain cell cultures.

We recently proposed the involvement of diffusible modulators in signalling astrocytes to increase glial cell line-derived neurotrophic factor (GDNF) expression after selective dopaminergic injury by H2O2 or L-DOPA. Here we report that interleukin-1beta (IL-1beta) is involved in this crosstalk between injured neurons and astrocytes. IL-1beta was detected only in the media from challenged neuron-glia cultures. Exogenous IL-1beta did not change GDNF protein levels in astrocyte cultures, and diminished GDNF levels in neuron-glia cultures. This decrease was not due to cell loss, as assessed by the MTT assay and immunocytochemistry. Neither H2O2 nor L-DOPA induced microglia proliferation or appeared to change its activation state. The IL-1 receptor antagonist (IL-1ra) prevented GDNF up-regulation in challenged cultures, showing that IL-1beta is involved in the signalling between injured neurons and astrocytes. Since IL-1ra decreased the number of dopaminergic neurons in H2O2-treated cultures, we propose that IL-1 has a neuroprotective role in this system involving GDNF up-regulation.

GDNF modulates HO-1 expression in substantia nigra postnatal cell cultures.

Heme oxygenase-1 (HO-1) has been strongly highlighted because of its induction in many cell types by toxic stimuli, including oxidative stress. The intense HO-1 immunostaining in the substantia nigra of Parkinson disease (PD) patients suggests its involvement in the pathogenesis of this neurodegenerative disease. In this work we investigated HO-1 expression in rat substantia nigra postnatal cell cultures under conditions mimicking dopamine toxicity and its modulation by glial cell line-derived neurotrophic factor (GDNF), a potent neuroprotective factor for dopaminergic neurons. In neuron-glia cultures, we found that H2O2, a product of dopamine metabolism, or l-3,4-dihydroxyphenylalanine (L-DOPA), the dopamine precursor used in the therapy of PD, induced a fast up-regulation of HO-1 mRNA and protein levels, followed by a secondary down-regulation. H2O2 and L-DOPA also increased HO-1 expression in astrocyte cultures, but with a delayed time course in H2O2-treated cultures. HO-1 expression was decreased in neuron-glia cultures under conditions under which GDNF up-regulation was observed. Because exogenously applied GDNF prevented HO-1 up-regulation in cultures treated with H2O2 or l-DOPA, and antibody neutralization of GDNF prevented the secondary HO-1 down-regulation observed in neuron-glia cultures, we propose that GDNF negatively modulates HO-1 expression induced by oxidative stress. To our knowledge, this is the first report showing the modulation of HO-1 expression by GDNF.

[Intrathecal injection of muscarinic receptors or GDNF antisense oligonucleotides inhibits the increase of c-Fos expression in locus coeruleus of morphine-withdrawal rats].

The antisense approach and immunohistochemistry were used to study the effects of different muscarinic receptor (M) subtypes and glial cell derived neurotrophic factor (GDNF) on the scores of morphine-withdrawal syndrome and the expression of c-Fos in locus coeruleus (LC). Intrathecal injection of M2 receptor antisense oligonucleotides (M2AS-oligo) or GDNF antisense oligonucleotides (GDNFAS-oligo) decreased the scores of morphine withdrawal syndrome. The expression of c-Fos positive neurons in the LC increased in morphine-dependent rats and increased to a greater extent after the injection of naloxone (4mg/kg, ip) in morphine dependent rats. Intrathecal injection of M2AS-oligo or GDNFAS-oligo inhibited the increase of c-Fos expression in LC during morphine withdrawal, but there was no effect in case of M1AS-oligo. The results suggest that M2 receptor of spinal cord mediates the neural activation of LC during morphine withdrawal. And the interaction between neurons and glial cells may be involved in the ascending activation process.

17beta-estradiol rescues spinal motoneurons from AMPA-induced toxicity: a role for glial cells.

The ability of astrocytes to mediate 17beta-estradiol neuroprotection of spinal motoneurons challenged with AMPA has been evaluated in a co-culture system in which pure motoneurons were pulsed with 20 microM AMPA and then transferred onto an astrocyte layer pretreated for 24 h with 10 nM 17beta-estradiol. Under these conditions, AMPA toxicity was reverted, an effect that was likely related to increased production and release of GDNF, as shown by RT-PCR, Western blot analysis and ELISA assay. In addition, treatment with GDNF during the 24 h that followed the AMPA pulse produced a similar neuroprotective effect, whereas addition of a neutralizing anti-GDNF antibody prevented neuroprotection. These data suggest a role for astrocytes in the neuroprotective effect of 17beta-estradiol against spinal motoneuron death and find strong support in the marked up-regulation of estrogen receptor alpha found in spinal astrocytes of amyotrophic lateral sclerosis patients.