Astrocyte-secreted GDNF and glutathione antioxidant ...

Neurobiology of Disease 33 (2009) 405–414

Contents lists available at ScienceDirect

Neurobiology of Disease j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / y n b d i

Astrocyte-secreted GDNF and glutathione antioxidant system protect neurons against 6OHDA cytotoxicity Jagdeep K. Sandhu a,⁎,1, Mossa Gardaneh a,1, Rafal Iwasiow a, Patricia Lanthier a, Sandhya Gangaraju a, Maria Ribecco-Lutkiewicz a, Roger Tremblay a, Kazutoshi Kiuchi b, Marianna Sikorska a,⁎ a b

Neurogenesis and Brain Repair Group, M54, Institute for Biological Sciences, National Research Council Canada, 1200 Montreal Road, Ottawa, Ontario, Canada K1A 0R6 Department of Biomolecular Science, Faculty of Engineering, Gifu University, 1-1 Yanagido, 501-1193, Gifu, Japan

a r t i c l e

i n f o

Article history: Received 24 July 2008 Revised 19 November 2008 Accepted 20 November 2008 Available online 11 December 2008 Keywords: Neuroprotection Neuro-regeneration Neurotrophic factor Signal transduction Thiol antioxidant Coculture Viral vectors Lentiviral vector Oxidative stress Gene therapy

a b s t r a c t In recent years, GDNF has emerged as a protective and restorative agent in several models of neurodegeneration; however, the exact molecular mechanisms responsible for these effects are not yet fully understood. Here we examined the effects of astrocytes secreting GDNF on neurons subjected to 6OHDA toxicity using in vitro neuron-astroglia co-cultures. Astrocytes were transduced with lentiviral vectors carrying the GDNF gene under the control of either human glial fibrillary acidic protein or cytomegalovirus promoters. The overexpression of GDNF, regardless of the promoter employed, had no obvious adverse effects on astroglia and the engineered cells stably produced and secreted GDNF for extended periods of time (≥3 weeks). These astrocytes very effectively protected neurons against 6OHDA, in both mouse and human co-culture systems. The neuroprotective effects were mediated not only by GDNF, but also by the antioxidant GSH since its depletion reduced the level of GDNF protection. Furthermore, neurons and astrocytes expressed different components of GDNF signaling complex, suggesting that they might utilize separate pathways to mediate autocrine and paracrine effects of GDNF. © 2008 Elsevier Inc. All rights reserved.

Introduction Glial cell-line derived neurotrophic factor (GDNF) was originally isolated from a rat B49 glioblastoma cell-line supernatant and subsequently shown to be a potent survival factor for midbrain dopaminergic neurons (Lin et al., 1993). Since then GDNF has also been shown to promote the survival and differentiation of many other types of neuronal populations, such as spinal and central motor neurons, brainstem noradrenergic neurons, basal forebrain cholinergic neurons, enteric neurons, cerebellar Purkinje cells and sensory and sympathetic neurons (Oppenheim et al., 1995). GDNF is a member of the transforming growth factor-β superfamily and its neurotrophic action is mediated by a unique multicomponent receptor system consisting of the GDNF-family of receptors (GFRα1q-4), and Ret tyrosine kinase encoded by the c-ret proto-oncogene (Airaksinen and Saarma, 2002; Sariola and Saarma,

Abbreviations: 6OHDA, 6 hydroxydopamine; GFP, green fluorescent protein; GDNF, glial cell line-derived neurotrophic factor; GSH, glutathione. ⁎ Corresponding authors. Fax: +1 613 941 4475. E-mail addresses: [email protected] (J.K. Sandhu), [email protected] (M. Sikorska). 1 Equal contributions. Available online on ScienceDirect (www.sciencedirect.com). 0969-9961/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.nbd.2008.11.016

2003). GDNF can also activate intracellular signaling pathways through GFRα1q-4 receptor in the absence of Ret (Poteryaev et al., 1999; Trupp et al., 1999; Saarma and Sariola, 1999). Numerous studies, carried out in animal models of neurodegeneration, demonstrate that exogenously administered GDNF supports long-term neuronal survival and contribute to behavioral improvements. Thus, it protects and repairs dopaminergic neurons, which degenerate in Parkinson's disease (Kordower et al., 2000; Eslamboli, 2005; Sherer et al., 2006; Gill et al., 2003), striatal projection neurons and cortical neurons, which degenerate in Huntington's disease (McBride et al., 2003; McBride et al., 2006) and motor neurons, which die in amyotrophic lateral sclerosis (Wang et al., 2002; Bohn, 2004; Klein et al., 2005). Administration of GDNF to areas of ischemic brain injury also limits cerebral infarction, promotes neurogenesis and reduces damage to motor functions in stroke (Horita et al., 2006; Kobayashi et al., 2006). A number of GDNF delivery methods into the CNS have been tested in animal models of neurodegeneration, including a direct tissue infusion, adenoviral and lentiviral infections or engineered cell lines such as fibroblasts, myoblasts, CNS progenitors and glia (Akerud et al., 2001; Cunningham and Su, 2002; Capowski et al., 2007). These studies continue to support the therapeutic value of GDNF, but also reveal several shortcomings and concerns. Thus, a continuous GDNF delivery seems to induce a downregulation of tyrosine hydroxylase protein

406

J.K. Sandhu et al. / Neurobiology of Disease 33 (2009) 405–414

in the preserved striatal dopamine terminals and aberrant sprouting of tyrosine hydroxylase neurons in globus pallidus, entopeduncular nucleus and substantia nigra (Georgievska et al., 2002; Rosenblad et al., 2003). Furthermore, adeno-associated viral delivery of GDNF reported an increase in neuronal death and failure to ameliorate behavioral deficits after stroke in rats (Arvidsson et al., 2003). Therefore, the CNS delivery of GDNF still remains problematic and requires further studies. Astrocytes, which perform many functions critical for neuronal survival, are very promising candidates for the GDNF delivery (Wilson, 1997; Dringen and Hirrlinger, 2003; Byrd et al., 2004). These cells can be easily engineered to produce and secrete the most beneficial doses of this neurotrophin (Cunningham and Su, 2002; Parsadanian et al., 2006; Pertusa et al., 2007). However, several mechanistic questions regarding the effects of GDNF overexpression on both neurons and astrocytes need to be critically assessed. Accordingly, we have developed lentiviral vectors carrying the GDNF transgene under the control of either cell-type specific weak promoter of human glial fibrillary acidic protein or non-specific strong cytomegalovirus promoter. The vectors were used to transduce astrocytic cells, which, subsequently, were tested for neuroprotection against 6OHDA in both mouse and human neuron/astroglia cocultures. Here we report that the overexpression and long term secretion of GDNF did not alter astrocytic morphology or their proliferation rate, instead, GDNF was clearly neuroprotective. Furthermore, this neuroprotective effect was synergistic with the GSH antioxidant system. The data also revealed that neurons and astrocytes expressed and, most likely, utilized different components of the GDNF signaling complex as neurons expressed GFRA1α and Ret tyrosine kinase co-receptor, whereas astrocytes expressed predominantly GFRA1α. Materials and methods Construction of lentiviral constructs Lentiviral plasmids pWPT-GFP (transfer vector), pMD-G (envelope vector) and pR8.74 (packaging vector) were kindly provided by Dr. D. Trono (University of Geneva, Switzerland, http://www.tronolab.com). The plasmid pcDNA3.1/Zeo(+)-mGDNF(Kozak) was supplied by Dr. K. Kiuchi (GIFU, Japan) (Matsushita et al., 1997). The hGFAP promoter was a kind gift from Dr. J. He (Indiana University School of Medicine, Indianapolis, USA). Lentiviral constructs encoding the glial cellderived neurotrophic factor (mGDNF) or green fluorescent protein (GFP) driven by the 5′LTR and glial fibrillary acidic protein (GFAPp) or cytomegalovirus (CMVp) transcriptional promoters were constructed and used in our study (Fig. 1).

Cell culture Mouse N2a cell culture Mouse neuroblastoma Neuro-2a cells (N2a) (ATCC, CCL-131) were maintained as monolayer cultures in Dulbecco's modified minimal essential medium with high glucose (DMEM, Invitrogen, Burlington, ON) supplemented with 10% fetal bovine serum (FBS, Wisent, SaintJean, QC), L-glutamine and pyridoxine hydrochloride at 37 °C in a humidified atmosphere containing 95% air and 5% CO2. Mouse ventral mesencephalon astrocyte cultures Primary cultures of ventral mesencephalon astrocytes (VM-astrocytes) were prepared from post-natal day 4 CD-1 mice (Charles River, St. Constant, QC) in accordance with the Canadian Council on Animal Care and the procedures approved by the institutional Animal Care Committee. Briefly, brains were removed and the ventral mesencephalon tissue was dissected, mechanically dissociated and cultured in DMEM containing 10% FBS at 6 × 105 cells/mL in poly-L-lysine coated plates. Cultures were fed by replacing half medium with fresh culture medium every 3 days. Cultures were maintained in DMEM containing 10% FBS and used for experiments at 10–21 days in vitro. Cocultures of VM-astrocytes and N2a cells Cocultures were prepared by seeding 4 × 104 mouse VM-astrocytes in a 12-well dish and allowed to attach for 3 days. N2a cells at 7.5 × 104 were then seeded on top of VM-astrocytes and allowed to attach for 24 h. Differentiation of N2a cells into neuron-like cells was initiated by changing the culture media to 0.5% FBS for 3 days and then to DMEM containing 10% FBS. Human NT2 cell culture NT2-D1 progenitor cells (Stratagene, La Jolla, CA, USA) were seeded at a density of 2 × 106 cells per T75 cm2 flask and treated with 10 μM all-trans-retinoic acid (RA, Sigma, Oakville, ON) for 4 weeks. Fresh RA and high glucose DMEM medium supplemented with 10% FBS were replenished every 2 days. Cells were then harvested by trypsinization and transferred into T175 cm2 flasks for 24–48 h in DMEM in the absence of RA. Cocultures of NT2-derived astrocytes and neurons were prepared as described previously (Byrd et al., 2004; Sandhu et al., 2002). Virus production and transduction of astrocytes with the GFP or GDNF lentiviruses HEK293T cells (2 × 106) were seeded in 10 cm round plastic dishes 24 h before transfection. Transient transfection was carried out using standard calcium phosphate precipitation method. Transfer vector (10 μg), envelope and packaging vectors (5 μg each) were used for cotransfection of 293T cells to generate recombinant lentiviruses. Six hours post-transfection, medium was removed, cells were washed with phosphate buffered saline (PBS) and fed with fresh medium supplemented with 10% FBS and incubated at 37°C, 5% CO2. Forty eight hours post transfection the supernatant was collected and centrifuged at 3000 rpm for 10 min to remove cell debris, filtered using 0.45 μm filters (Millipore) and concentrated using Amicon Ultra-15 spin columns (100,000 mol. wt. cut off, Millipore). The concentrated virus was divided into fresh eppendorf tubes and polybrene (8 μg/mL) was added and incubated for 10 min. The virus was then applied to astrocyte cultures and media was changed 24 h later. GDNF immunoassay

Fig. 1. Schematic representation of lentiviral constructs encoding the glial cell-derived neurotrophic factor (mGDNF) or green fluorescent protein (GFP) driven by either glial fibrillary acidic protein (GFAPp; GFAP-GFP and GFAP-GDNF) or cytomegalovirus (CMVp, CMV-GFP and CMV-GDNF) transcriptional promoters.

GDNF protein levels were determined using a GDNF E® max Immunoassay system according to the supplier's protocol (Promega, Madison, WI). Briefly, ELISA plates (Maxisorp 96-well flat-bottomed dishes, Nalge Nunc International) were coated with the capture


antibody (monoclonal antibody to human GDNF) diluted in carbonate coating buffer, pH 8.2 and incubated overnight at 4 °C. Wells were blocked for 1 h at room temperature with 1× blocking buffer (200 μL/ well). GDNF standards ranging from 0–1000 pg/100 μL were prepared using recombinant human GDNF and samples (100 μL, dilutions ranging from 5-fold to 1000-fold) were applied to the wells. All samples were incubated with shaking for 6 h at room temperature and then washed with TBS-T (20 mM Tris–HCl, pH 7.6, 150 mM NaCl, 0.05% (v/v) Tween 20). The bound GDNF was detected by chicken polyclonal antibody to human GDNF incubated overnight at 4 °C. Following washes with TBS-T, horseradish peroxidase-conjugated antichicken IgY antibody was added to the plates and incubated with shaking at room temperature for 2 h. The plates were again washed with TBS-T, and 100 μL of the enzyme substrate (Tetramethylbenzidine One solution) was added. The plates were incubated for 15 min at room temperature in the dark and the reaction was stopped by the addition of 100 μL 1 N HCl per well. The absorbance was measured at 450 nm and the amount of GDNF was calculated from the standard curve in the linear range. The detection limit of the assay is ≥3 pg/mL of GDNF. Western blotting Total proteins were extracted from VM-astrocytes in RIPA buffer (50 mM Tris–HCL, pH 7.4, 150 mM NaCl, 2 mM EDTA, 1% Triton X-100, 0.1% SDS, 1% sodium deoxycholate) containing 1 mM PhenylMethaneSulfonyl Fluoride and 1× protease inhibitor cocktail. Proteins were separated by 12% SDS-PAGE and transferred to nitrocellulose membranes. The membranes were blocked with 5% skim milk in TBST (10 mM Tris–HCl, pH 8.5, 150 mM NaCl, 0.1% Tween-20), probed with rabbit polyclonal anti-GDNF antibody (1:500, Santa Cruz Biotechnology Inc.) and detected using horseradish peroxidaseconjugated anti-rabbit IgG (1:5000, Sigma). Immunoreactive bands were detected with the ECL plus western blotting detection system (Amersham Pharmacia Biotech, Quebec). Treatment of cells with 6 hydroxydopamine A stock solution of 6 hydroxydopamine (5 mM, 6OHDA, Sigma Chemical Co.) was prepared as described previously (Ding et al., 2004). Briefly, a metal chelator, diethylenetriaminepentaacetic acid (10 mM, Sigma) and ascorbic acid (0.15%, Sigma) was dissolved in serum free medium, flushed with nitrogen gas on ice for 10 min and 6OHDA was added at a concentration of 5 mM. Cells were treated for 15 min, washed and further incubated for 16 h at 37 °C. Treatment of cells with buthionine sulfoximine N2a or VM-astrocytes cultures were treated with L-buthionine-(S, R)-sulfoximine (BSO, ICN Biomedicals Inc.) as described previously (Byrd et al., 2004). For GSH depletion, cells were incubated with 50 μM BSO for 24 h. Cell viability assays MTT reduction assay Cell viability was determined by the level of cellular MTT (3-(4,5dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide, MTT cell proliferation kit, ATTC, USA) reduced by metabolically active cells as per manufacturer's protocol. Briefly, cocultures of N2a and VMastrocytes were treated with 6OHDA and recovered as described above. Twenty five microliter of MTT reagent was added and further incubated at 37°C for 4 h. At the end of the incubation period, 100 μL of detergent reagent was added to each well and further incubated for 2 h and absorbance was measured at 570 nm. Cell viability was expressed as a percentage of the untreated control culture.

407

Hoechst staining Following treatments, cells were fixed with 4% paraformaldehyde, washed and mounted in DAKO mounting medium (DAKO Diagnostics Canada Inc.) spiked with 5 μg/mL Hoechst 33258. The percentage of cells with condensed and shrunken nuclei was determined using a 20× objective under a Carl Zeiss Axiovert 200M microscope. Approximately 800–1500 cells were counted in 5–6 random fields from 2 separate experiments. Immunocytochemistry Cultures grown on glass coverslips were fixed with Genofix™ (DNA Genotek Inc., Ottawa, ON) for 10 min at room temperature, rinsed with PBS and blocked for 30 min with universal blocking solution (DAKO Diagnostics Canada Inc., Mississauga, ON). Excess liquid was drained and cells were incubated for 1 h at room temperature with the following primary antibodies: anti-glial fibrillary acidic protein, 1:100 (GFAP, mouse monoclonal clone GA-5, Neomarkers); anti-βIII tubulin, 1:20 (mouse monoclonal clone Tuji, Dr. Dave Brown, University of Ottawa) and anti-GFP, 1:100 (rabbit polyclonal, Chemicon International). Cells were washed with PBS and incubated for 45 min at room temperature with Rhodamine Red™-conjugated goat anti-mouse IgG, Alexa 488-conjugated goat anti-mouse or anti-rabbit IgG (1:600, Molecular Probes Inc.). Negative controls with omission of primary antibodies were included. The images were captured on a Carl Zeiss Axiovert 200M microscope. Real-time PCR Total RNA was extracted using Trireagent (Molecular Research Center Inc., Cincinnati, OH), according to the manufacturer's instructions. RNA was reverse transcribed using Superscript II reverse transcriptase enzyme (Invitrogen- Burlington, ON) and AncT (5′ T20VN3′) primers according to the manufacturer's instructions. Following cDNA synthesis, RNA templates in the RT reaction were degraded, cDNA was purified using QIAquick PCR purification columns (Qiagen- Mississauga, ON) and quantitated using a Quant-iT™ OliGreen® ssDNA quantitation kit (Molecular Probes, Invitrogen, Burlington, ON). Two and a half nanograms of cDNA per sample, in triplicates, were used for real time PCR, performed using primer sets specific for the GDNF receptor genes (Table 1) and SYBR® Green PCR Master Mix (Applied Biosystems, Foster City, CA) in the ABI PRISM® 7000 Sequence Detection System (Applied Biosystems). The primers were designed using Primer Express (Applied Biosystems). Changes in gene expression were assessed using a comparative Ct (2 −ΔΔCt) method to calculate relative fold differences between GFRA1 (alpha plus beta isoforms; for human cells) and Gfra1 (for mouse cells) and other GDNF receptors. Measurement of glutathione content After washing with ice-cold PBS, approximately 1 × 106 cells were harvested by centrifugation at 194 ×g for 5 min. The cell pellets were re-suspended in 500 μL of cold deoxygenated PBS and disrupted by ultrasonication. Proteins were precipitated with ice-cold perchloric acid (at a final concentration of 5%), and supernatants and pellets were collected by centrifugation at 14,000 ×g at 4°C for 10 min. The supernatants were transferred to fresh eppendorf tubes, frozen immediately and stored at − 80 °C until analyzed. Pellets were dissolved in 0.1 M NaOH and the protein content was determined using BioRad reagent with bovine serum albumin as a standard. Supernatants were injected directly into the HPLC (Beckman System Gold) equipped with a Synergi 4 μL Hydro-RP, 150 × 4.6 mm column (Phenomenex, Torrence, CA) preceded by a SecurityGuard C18 aqua guard cartridge. The mobile phase consisted of a mixture of 20 mM KH2PO4 and 1% acetonitrile (v/v), pH 2.7 at room temperature.

408


Table 1 Primer sequences for real-time PCR Human gene

Primer sequence

GDNF

F: 5′ TGGGTCTGGGCTATGAAACC 3′ R: 5′ GTCTCAGCTGCATCGCAAGA 3′ F: 5′ AACCAGTTAACAGCAGATTGTCAGA 3′ R: 5′ ATGTGCTCCACTTGCTGAAAAA 3′ F: 5′ TGCTCCTATGAAGAGAGGGAGAA 3′ R: 5′ AAATCCGCAAGGCGAGATCT 3′ F: 5′ GCGCCAAGAGCAACCATT 3′ R: 5′ GGAGCGCAGCTTCTTGCA 3′ F: 5′ GTGAAATGCTGGAAGGGTTCTT 3′ R: 5′ AACGCATCTTAGCTGCAATGG 3′ F: 5′ AGCATATAACCCCCTAAGGAATCC 3′ R: 5′ AATGGTGGGTAATGCCCAGTT 3′ F: 5′ CGATGTTGTGGAGACCCAAGA 3′ R: 5′ GCTCGTGTCCCCCAACAAT 3′

GFRA1 alpha GFRA1 total (aplha and beta) GFRA2 GFRA3 GFRA4 RET Mouse gene

Primer sequence

Gdnf

F: 5′ TCGGCCGAGACAATGTATGA 3′ R: 5′ CAACATGCCTGGCCTACTTTG 3′ F: 5′ CCAGCGGGAACTCCTTTGT 3′ R: 5′ GCCCTGTAGCAGTTCTTCAACA 3′ F: 5′ CACCACCTGCACATCTATCCA 3′ R: 5′ GAGCTCTGTGAAACACATGCTTAAC 3′ F: 5′ CAGACCCACTGTCATCCTATGGA 3′ R: 5′ CAGGTATGCCCGCAGACAT 3′ F: 5′ GGCAGAAACAGTCCTTGTTTTGT 3′ R: 5′ GGAGAGCCAGGGCAGTGA 3′ F: 5′ TGACCATGGGTGACCTCATCT 3′ R: 5′ TACAAGCTTCATTTCTGCCAAGTACT 3′

Gfra1 Gfra 2 Gfra 3 Gfra 4 Ret

A constant flow rate of 1.0 mL/min was applied. GSH detection was achieved with a UV detector set at 210 nm. Reduced Glutathione (GSH, Sigma) was used as a standard; 0.2 μg was injected into the column and eluted as above with a retention time of 3.217 min. Data analysis was performed using 32 Karat software (Beckman Coulter). Statistical analysis Data are represented as the mean ± SEM of 3–4 separate experiments carried out in triplicate. Differences between two groups were

analyzed by unpaired Student's t-test. Differences between three or more groups were analyzed by one-way analysis of variance (ANOVA), followed by post-hoc Dunnett's or Newman–Keuls multiple-comparisons test in GraphPad Prism version 4.0. A value of P b 0.05 (#) was considered to be statistically significant and P b 0.01 (⁎⁎ or ##) and 0.001 (⁎⁎⁎, ### or †††) as highly statistically significant. Results Transgene expression in astrocytes We have designed and constructed lentiviral vectors (Fig. 1) carrying either GFP or GDNF under the control of two different promoters: human glial fibrillary acidic protein, hGFAPp (GFAP-GFP and GFAP-GDNF) and the cytomegalovirus, CMVp (CMV-GFP and CMV-GDNF), respectively. The abilities of these vectors to deliver the transgenes were tested in primary mouse ventral mesencephalon astrocytes (VM-astrocytes). As shown in Fig. 2, VM-astrocyte cultures were efficiently transduced with both GFAP-GFP and CMV-GFP lentiviral constructs, but the GFP signal was much stronger in astrocytes transduced with the CMV-GFP (Fig. 2D), than with the GFAP-GFP (Fig. 2B) viral vector, consistent with the properties of these promoters. Overexpression and secretion of GDNF The production and secretion of GDNF was tested in VM-astrocytes transduced with the GFAP-GDNF and CMV-GDNF vectors (Fig. 3A). The cells infected with the GFAP-GDNF construct secreted 57.3 ± 9.3 pg/mL of GDNF after 48 h, whereas cells transduced with the CMV-GDNF secreted approximately a 40-fold higher level of GDNF (Fig. 3A; P b 0.001, Newman–Keuls multiple-comparisons test). Western blot analysis showed the increased presence of GDNF oligomers in the transduced VM-astrocytes, in addition to a single band of 17 kDa, seen also in control astrocytes (Fig. 3B). The secretion of GDNF by the transduced astrocytes was also tested in the presence of neurons in both mouse and human co-culture models (Fig. 3C). The mouse model consisted of the GDNF producing primary VM-astrocytes and N2a neuroblastoma derived neurons

Fig. 2. GFP expression driven by the GFAPp and CMVp promoters. Primary mouse VM-astrocytes, grown on glass coverslips, were transduced with lentiviral vectors, GFAP-GFP (A and B) or CMV-GFP (C and D). The cells were then fixed and labeled with anti-GFAP (red; A) or anti-GFP (green; B and D) antibodies. Shown is also a phase-contrast image (C). Bars = 20 μm.


409

culture media was ~4-fold higher than its intracellular accumulation (data not shown). Cytotoxicity of 6OHDA and GNDF neuroprotection in mouse cells The neuroprotective effects of astrocytic GDNF delivery against 6OHDA-induced cytotoxicity were subsequently tested in the neuron-

Fig. 3. Production of GDNF by transduced VM-astrocytes. (A) Secretion of GDNF into the media. Astrocytes were transduced with GFAP-GDNF (white bar) or CMV-GDNF (hatched bar) and the GDNF concentration in the media was measured by ELISA, as described in the Materials and methods. Statistical differences between astrocytes transduced with GFAP-GDNF or GFAP-GFP and CMV-GDNF or CMV-GFP are indicated by ⁎⁎⁎ (P b 0.001). N.D. =not detected. (B) Western blotting. Cells transduced with GFAP-GDNF, CMV-GDNF or CMV-GFP vectors were lysed in RIPA buffer and protein samples, separated on 12% SDS-PAGE and electrotransferred onto nitrocellulose membranes, were immunoblotted with anti-GDNF antibody, as described in the Material and Methods. (C) Secretion of GDNF into the media by the neuron/astroglia co-cultures. The concentrations of GDNF were measured by ELISA in the media collected from the co-cultures of N2a cells and VM-astrocytes transduced with GFAP-GDNF (white bar) and CMV-GDNF (hatched bar) vectors, as well as, from the co-cultures of NT2 neurons and NT2-astrocytes transduced with the same vectors (GFAP-GDNF – light grey bar and CMV-GDNF – black bar). Statistical differences between GDNF produced by VM-astrocytes or NT2-astrocytes transduced with GFAP-GDNF or CMV-GDNF are indicated by ⁎⁎⁎ (P b 0.001). N.D. = not detected.

(Fig. 4C, b). The human model was represented by the GDNF transduced NT2-astrocytes and NT2-neurons (Fig. 5B, b). Interestingly, the GDNF-secretion into the media was lower (~ 2–5 fold) when the transduced astrocytes were grown in the presence of neurons (Fig. 3C). Thus, in the mouse co-culture system the GDNF concentration in the media was 36 ± 2.3 pg/mL from the GFAP-GDNF vector and 464 ± 50 pg/mL from the CMV-GDNF vector (Fig. 3C, GFAP-GDNF versus CMV-GDNF, P = 0.004, unpaired Student's t-test) as compared to the levels produced by the astrocytic cultures alone (Fig. 3A). Similarly, NT2-astrocytes produced 40.5 ± 6.6 pg/mL of GDNF from GFAP-GDNF construct and 529 ± 59 pg/mL from CMV-GDNF (Fig. 3C, GFAP-GDNF versus CMV-GDNF, P = 0.0002, unpaired Student's t-test). A consistent and stable secretion of GDNF was observed for at least 3 weeks posttransduction (data not shown), suggesting that no downregulation of the transgene expression occurred under in vitro culture conditions. The amount of GDNF secreted by the transduced astrocytes into the

Fig. 4. Toxic effects of 6OHDA on mouse cells. (A) Viability of 6OHDA-treated N2a cells. N2a cells were differentiated for 3 days in low serum conditions and were subsequently treated with increasing concentrations of 6OHDA. Cell viability was determined by the MTT assay as described in the Materials and methods. Statistical differences between vehicle and 6OHDA-treated N2a cells are indicated by ⁎⁎⁎ (P b 0.001). Ctrl = viability of untreated cells. (B) Photomicrographs of differentiated N2a cells (stained with anti-βIII tubulin (A, C) or tyrosine hydroxylase antibodies) and treated with vehicle (A, B) or 6OHDA (C). Bars = 20 μm. (C) Photomicrographs of VM-astrocytes grown in the absence (A) or presence of N2a cells (B) and treated with 6OHDA. Untreated coculture of N2a cells and VM-astrocytes is also shown (C). In panel B, arrow points to normal astrocyte nuclei and arrow head indicates apoptotic neuronal nuclei. Bars = 20 μm. (D) Viability of 6OHDA-treated N2a/VM-astrocyte co-cultures. Differentiated N2a cells were cocultured with the increasing numbers of VM-astrocytes and treated with 100 μM 6OHDA, as described in the Materials and methods. Statistically significant differences between the viability of N2a cells treated with 100 μM 6OHDA in the absence or presence of VM-astrocytes are indicated by ⁎⁎ (P b 0.01). Ctrl = viability of untreated cells.

410


Fig. 5. Neuroprotective effects of GDNF on 6OHDA-treated N2a cells. The co-cultures of differentiated N2a cells and VM-astrocytes transduced with GFAP-GFP (white bar), CMV-GFP (dark bar), GFAP-GDNF (light bar) or CMV-GDNF (hatched bar) were treated with 100 or 400 μM 6OHDA and cell viability was determined by the MTT assay. Statistical differences between cell survival in control co-cultures (GFAP-GFP OR CMV-GFP) versus GFAP-GDNF or CMV-GDNF transduced VM-astrocytes are shown as ⁎⁎ (P b 0.01) or ⁎⁎⁎ (P b 0.001) and between GFAP-GDNF versus CMV-GDNF transduced VM-astrocytes are shown as # (P b 0.05) or ## (P b 0.01). Ctrl = viability of untreated cells.

astroglia co-cultures. Mouse N2a cells were differentiated into neurons by a standard method of lowering serum in the growth media. The differentiated cells extended neurites and expressed the

neuronal marker, βIII tubulin (Fig. 4B, a) and also expressed the dopamine cell marker, tyrosine hydroxylase (Fig. 4B, b). These cells were very sensitive to 6OHDA treatment. A statistically significant decrease in cell viability was observed at all the tested concentrations of 6OHDA (Fig. 4A; ANOVA, P b 0.0001). For example, 50% of cells lost viability in response to 25 μM 6OHDA and at concentrations higher than 50 μM, only 5–10% of the cells remained viable 24 h later. Microscopic examination of the 6OHDA treated cultures showed characteristic changes of cell death including the retraction and loss of neuronal processes and nuclear condensation (Fig. 4B, b). The primary VM-astrocytes, on the other hand, were insensitive to 6OHDA; no phenotypic changes or loss of viability were observed in response to the same concentrations of the drug (Fig. 4C, a). The astrocytes are equipped with a robust antioxidant system and are known to protect neurons during oxidative stress (Peuchen et al., 1997; Takuma et al., 2004; Trendelenburg and Dirnagl, 2005). Thus, prior to testing the effects of GDNF in the co-culture model, the effects of astrocytes themselves on the 6OHDA treated N2a neurons had to be established. Indeed, the viability of N2a neurons exposed to 100 μM 6OHDA in the presence of increasing numbers of astrocytes was significantly enhanced and correlated with the number of astrocytes in the culture (Fig. 4D). For example, in the presence of 2.5 × 104 VMastrocytes a 20% enhancement of the cell survival was seen (P b 0.01; Dunnett's multiple-comparisons test) and near complete protection of neuronal cells from the effects of 100 μM 6OHDA was achieved when

Fig. 6. Toxic effects of 6OHDA on human NT2 cells. (A) Viability of 6OHDA-treated NT2-neurons. NT2-neurons were treated with increasing concentrations of 6OHDA and cell viability was determined by the MTT assay. Statistically significant differences between untreated and 6OHDA-treated cultures are shown as ⁎⁎⁎ (P b 0.001). Sister cultures, grown on glass coverslips, were treated with 6OHDA and viewed under a microscope. Inset show neuronal nuclei with condensed chromatin and apoptotic morphology. Ctrl = viability of untreated cells. Bar = 30 μm. (B) Viability of 6OHDA-treated NT2-astrocytes. NT2-astrocytes were treated with increasing concentrations of 6OHDA and cell viability was determined by the MTT assay. Sister cultures, grown on glass coverslips, were treated with 6OHDA and viewed under a microscope. Ctrl = viability of untreated cells. Bar = 30 μm. (C) Viability of 6OHDAtreated NT2 neuron/astrocyte co-cultures. NT2-neurons, grown in the presence of astrocytes, were treated with increasing concentrations of 6OHDA and scored for viability. Statistically significant differences between untreated and 6OHDA-treated cultures are shown as ⁎⁎⁎ (P b 0.001). Arrow points to normal astrocyte nuclei and arrow head indicates normal neuronal nuclei. Ctrl = viability of untreated cells. Bar = 30 μm. (D) Neuroprotective effects of GDNF on 6OHDA-treated NT2 neuron/astrocyte co-cultures. NT2-neurons were co-cultured with astrocytes transduced with control CMV-GFP (black bar), GFAP-GFP (dark bar), GFAP-GDNF (light bar) or CMV-GDNF (hatched bar) vectors and were treated with 100 μM 6OHDA. Cell viability was determined by counting Hoechst stained nuclei with condensed chromatin and apoptotic morphology (Panel A, inset), as described in Materials and methods. Statistically significant differences between cell survival in control co-cultures with GFAP-GFP or CMV-GFP transduced astrocytes versus co-cultures containing GDNF overexpressing NT2-astrocytes (GFAP-GDNF or CMV-GDNF) are indicated by ⁎⁎⁎ (P b 0.001). Ctrl = viability of untreated cells.


411

the number of VM-astrocytes was increased up to 1.5 × 105 (Fig. 4D; P b 0.01, Dunnett's multiple-comparisons test). These results were used to establish the optimal ratio of neurons to astrocytes in the co-cultures that would allow assessing neuroprotective effects of GDNF against 6OHDA. Thus, the co-cultures consisting of 7.5 × 104 N2a neurons and 4 × 104 lentivirally transduced VM-astrocytes were set up and treated with the high concentrations of 6OHDA, i.e., 100 μM and 400 μM, which were totally lethal to N2a cells alone (Fig. 4A). At this ratio of N2a cells and control astrocytes infected with CMV-GFP, approximately 40–50% of N2a neurons lost viability following exposure to 100 μM and 400 μM 6OHDA, respectively (Fig. 5). Microscopic examination of these co-cultures showed that the dead or dying cells were neurons; no cell death was seen in the underlying astrocytes (data not shown). On this background, the protective effects of astrocytes producing GDNF were clearly observed. Thus, astrocytes transduced with the GFAP-GDNF or CMV-GDNF lentiviral vectors further enhanced the N2a cell survival following the exposure to both dosages of 6OHDA. The degree of protection was higher from the CMV-GDNF transduced astrocytes secreting higher level of GDNF than that from the GFAP-GDNF (Fig. 5B, P b 0.05 for 100 μM 6OHDA and P b 0.01 for 400 μM 6OHDA, Newman– Keuls multiple-comparisons test). Cytotoxicity of 6OHDA and GNDF neuroprotection in human cells The neuroprotective effects of GDNF secreting astrocytes against 6OHDA were also tested in human cells co-culture model (Fig. 6). In these co-cultures, the NT2-derived neurons established a network of interconnected processes and were found resting on the top of a monolayer of NT2 astrocytes (Fig. 6B, b). Here again, NT2 neurons cultured alone were very sensitive to 6OHDA. Their exposure to 50 μM resulted in a 50% loss of cell viability and the cells did not survive when treated with concentrations higher than 100 μM (Fig. 6A). Morphological examination of these neuronal cultures showed numerous dead cells with apoptotic nuclear morphology (Fig. 6A, micrograph insert). NT2 astrocytes, on the other hand, were resistant to the tested concentrations of 6OHDA (up to 250 μM); no loss of viability and no morphological changes indicative of cell death were seen in these cultures (Fig. 6B μM, micrograph). A higher survival rate was observed in neurons grown in the presence of astrocytes and subsequently challenged with 6OHDA (Fig. 6C). For example, 50% of the neurons survived when grown in the presence of astrocytes, while only b5% of the neurons survived when grown in the absence of astrocytes and challenged with 100 μM 6OHDA (Figs. 6A and C). Consistent with the data from the mouse co-culture system (Fig. 5), human NT2-neurons exposed to 6OHDA in the presence of GDNF secreting astrocytes (transduced with the GFAP-GDNF and CMV-GDNF vectors) also had a better survival rate (Fig. 6D; P b 0.001 for GFAP-GFP vs GFAP-GDNF and CMV-GFP vs CMV-GDNF, Newman–Keuls multiplecomparisons test). Morphological examination of 6OHDA-treated cocultures showed far fewer neuronal nuclei with condensed chromatin and here, again, no cell death was seen in the underlying astrocytes. Neuroprotective effects of GSH and GDNF Previous work from our laboratory, and others, has shown that the astrocyte GSH antioxidant system plays a crucial role in protecting neurons during oxidative stress (Dringen, 2000; Dringen and Hirrlinger, 2003; Byrd et al., 2004). Therefore, it was important to determine a contribution of GSH system to the neuroprotection delivered by GDNF-secreting astrocytes in these co-culture models. Accordingly, the effects of GSH depletion on the survival of N2a neurons were examined (Fig. 7). Glutathione was depleted selectively from cells by inhibiting its de novo synthesis using BSO, a selective inhibitor of γ-glutamyl cysteine synthase, a rate limiting enzyme in GSH synthesis (Meister, 1988). Thus, the exposure of N2a cells and

Fig. 7. Contribution of GSH to the GDNF neuroprotection. (A) Depletion of GSH by BSO treatment. Both cell types, N2a and VM-astrocytes, were incubated in the presence or absence of 50 μM BSO for 24 h. After the removal of BSO and replacement of the media (time 0) cell were harvested at the indicated time points and GSH concentrations were determined by HPLC method, as described in Materials and methods. (B) Neuroprotective effects of GSH and GDNF against 6OHDA. Co-cultures of differentiated N2a cells and VM-astrocytes transduced with CMV-GFP, GFAP-GDNF or CMV-GDNF were pretreated for 24 h with (hatched bar) or without (dark bar) 50 μM BSO and, subsequently, they were treated with 6OHDA, as described in the Materials and methods. Cell viability was determined by the MTT assay. Statistically significant differences between the cell survival following 6OHDA treatment of control co-cultures (CMV-GFP) versus cocultures containing the GDNF overexpressing astrocytes (GFAP-GDNF or CMV-GDNF) are indicated by ## (P b 0.01) or ### (P b 0.001), whereas, statistical differences between the BSO-treated and untreated co-cultures are indicated by ⁎⁎ (P b 0.01) or ⁎⁎⁎ (P b 0.001). In addition, the difference between the BSO treated co-cultures containing GFAP-GDNF and CMV-GDNF transduced astrocytes is indicated by ††† (P b 0.001).

VM-astrocytes to 50 μM BSO for 24 h resulted in approximately 6-fold reduction of GSH (from 25.18 to 4.35 pmol/μg protein) in N2a cells and approximately 10-fold reduction (from 40.04 to 3.75 pmol/μg protein) in VM-astrocytes. However, the fully reduced GSH level was maintained only for up to 6 h following the BSO removal (Fig. 7A). By 6 h, its concentration started to rebound in VM-astrocytes and by 24 h it fully recovered (not shown). Therefore, to separate out the contribution of the GSH antioxidant system from the neuroprotective properties of GDNF, the co-cultures were treated with the highest concentration of 6OHDA tested, i.e., 400 μM, in the absence or presence of BSO throughout the entire course of the experiment (Fig. 7B). No alteration in astrocyte morphology or cell death was seen in cultures treated with 6OHDA alone or with a combination of BSO and 6OHDA. Indeed, in every experimental setting, the depletion of GSH increased the sensitivity of N2a neurons to 6OHDA (Fig. 7B, grey vs. hatched bars) and allowed, clearly, to distinguish the neuroprotective effects of secreted GDNF. Thus, only approximately 20% of N2a cells survived when treated with 6OHDA in the presence of BSO and control astrocytes (Fig. 7B, CMV-GFP-hatched bar; P b 0.001, Newman–Keuls multiple-comparisons test), but between 40 and 60% remained alive in co-cultures with GFAP-GDNF or CMV-GDNF transduced astrocytes (P b 0.01, Newman–Keuls multiple-comparisons test).

412


Discussion

Fig. 8. Expression of GDNF signaling components in neurons and astrocytes. Total RNA was extracted from VM-astrocytes (A), NT2-astrocytes (B), differentiated N2a cells (C) and NT2-neurons (D) and analyzed by QPCR using primers specific for GDNF receptors (Table I), as described in the Materials and methods. The receptor expression levels were normalized to that of Gfra1 in mouse cells and GFRA1 in human cells. N.D. = not detected.

Expression of the GDNF signaling components GDNF utilizes a unique multicomponent receptor system consisting of GFRα1q-4 receptors and Ret tyrosine kinase co-receptor (Airaksinen and Saarma, 2002), the expression of which was analyzed by real time PCR in both mouse and human neurons and astrocytes (Fig. 8). Thus, primary mouse VM-astrocytes transcribed mainly Gfra1 message and negligible levels of Gfra4 and ret (relative to Gfra1), but did not express Gfra2 or Gfra3 (Fig. 8A). Similarly, NT2-astrocytes expressed predominantly GFRA1, a low level of GFRA2, but did not express Ret (Fig. 8B). The Ret message was present in both N2a neuroblastoma cells (Fig. 8C; at a considerably high level relative to Gfra1) and in NT2-neurons, which also expressed GFRA2 (Fig. 8D). Continuous exposure of neurons to astrocyte-secreted GDNF did not result in any morphological alteration or any change in the GDNF receptor expression (unpublished observations).

Numerous studies report potent neurotrophic effects of GDNF not only on dopaminergic neurons, but also on many other types of neuronal and non-neuronal cell populations. In animal models of neurodegeneration, GDNF consistently shows both neuroprotective and neuroregenerative effects when provided continuously either by means of viral vectors or through continuous pump-driven protein infusion (Kordower et al., 2000; McBride et al., 2006; Horita et al., 2006; Wang et al., 2002). However, despite potential benefits, especially for the treatment of PD, the CNS delivery methods of GDNF is not yet sufficiently robust and safe; hence, its clinical applications are still controversial. New studies continue to report the development of better and safer viral vectors or utilize cell lines suitable for the CNS grafting and delivery (Trono, 2000; Jakobsson and Lundberg, 2006; Wong et al., 2006; Kafri, 2004). Indeed, promising results have been obtained with cell grafts, including astrocytes, which have been successfully engineered to deliver GDNF (Zhao et al., 2004; Duan et al., 2005; Ericson et al., 2005; Pertusa et al., 2007; Galan-Rodriguez et al., 2008). However, the exact mechanisms by which neuroprotection and/or neurorestoration is observed in these studies are achieved are far less understood. Here we have examined the effects of astrocytes overexpressing and secreting GDNF on neurons subjected to 6OHDA-induced oxidative stress. We used in vitro neuron-astroglia co-cultures, which created conditions allowing the direct cell–cell contacts and communication. Both mouse and human astrocytes transduced with either GFAP or CMV promoter-driven vectors secreted biologically active GDNF and provided neuroprotection under oxidative stress. The obtained results were consistent between mouse and human cocultures. Although the astrocytes transduced with the CMV promoter secreted ~ 10 fold higher concentrations of GDNF in comparison to cells transduced with the GFAP promoter, a concomitant increase in neuronal viability was not observed (Figs. 5 and 6). This indicated that the higher amounts of GDNF are not always beneficial and, in fact, might account for the side-effects seen in both animal studies and human clinical trials (Kordower et al., 1999; Zurn et al., 2001; Gill et al., 2003; Arvidsson et al., 2003). The amount of GDNF secreted to the media by astrocytes maintained in the co-cultures was 2–5 fold lower compared to the astrocytic cultures alone (Fig. 3), suggesting that neurons might have utilized the secreted GDNF. This is consistent with the concept of the active metabolic interactions between neurons and astrocytes, which are essential for neuroprotection (Dringen et al., 2000). In our study, for example, 100 μM 6OHDA killed 90 % of N2a neuronal cells in the absence of astrocytes, whereas the same concentration of neurotoxin was nearly ineffective in the co-culture consisting of 2:1 ratio of astrocytes/neurons (Fig. 4). Thus, astrocytes per se had the survival-promoting effects probably because they can secrete several growth and neurotrophic factors and are equipped with a robust antioxidant system (Peuchen et al., 1997; Byrd et al., 2004; Sandhu et al., 2003). Although we did not measure the GSH contents in these cultures, it is plausible that the astrocytes counteracted the production of reactive oxygen species generated during exposure to 6OHDA by using their GSH antioxidant system (Wilson, 1997). We show previously that GSH depletion by BSO treatment compromises the interaction between neurons and astrocytes, which become less effective in protecting neurons (Byrd et al., 2004). It is known that astrocytes provide GSH precursors to the neurons to increase their GSH synthesis (Dringen and Hirrlinger, 2003). Published data shows a transcriptional upregulation of the GSH system in neurons grown in the presence of astrocytes in response to 6OHDA (Iwata-Ichikawa et al., 1999). In this study, we depleted GSH in the co-cultures secreting GDNF by the same BSO treatment and showed clearly that this resulted in the increased neuronal sensitivity to 6OHDA (Fig. 7). Therefore, the neuroprotection observed in the co-cultures of GDNF-secreting astrocytes and


neurons was the synergistic action of GDNF and GSH, which adds a further credence to the use of astroglia as neurotrophic factors delivery vehicles. The action of 6OHDA on astrocytes seems controversial at present. For example, a recent study of Raicevic et al. (2005) reports that 6OHDA is toxic to both rat and human astrocytes, whereas, we did not observe any toxic effects of similar concentration of 6OHDA on either mouse or human astrocytes. These differences are most likely derived from the experimental conditions applied in these studies. It is well known that 6OHDA undergoes spontaneous autooxidation to generate reactive oxygen species (ROS), such as hydrogen peroxide, superoxide anions and hydroxyl radical, hence if added directly to cell cultures may kill them due to the action of ROS. We have included a metal chelator diethylenetriaminepentaacetic acid and ascorbic acid in the preparation of 6OHDA (Ding et al., 2004), which limit its spontaneous autooxidation. The observed effects are then due to the intracellular action of the toxin subsequent to its uptake through the high affinity dopamine transporter. Clearly, in our experimental settings the astrocytes were insensitive to this action of 6OHDA. Astrocytes are the predominant cell type in the vicinity of glutamatergic and dopaminergic synapses, where they monitor and maintain low levels of glutamate and dopamine, respectively. Astrocytes express the dopamine (Karakaya et al., 2007) and glutamate (Sandhu et al., 2002; Sandhu et al., 2003) transporters, which allows the uptake of these neurotransmitters and specifically eliminate them from the extracellular space. Astrocytes are also equipped with a potent detoxification and GSH antioxidant system (Sandhu et al., 2003; Byrd et al., 2004), therefore they have the capacity to deal with higher levels of oxidative stress. Under our culture conditions, astrocytes seemed to perform these functions efficiently, hence, did not succumb to cell death. Interestingly, neurons and astrocytes expressed a different combination of the GDNF signaling complex components. Thus, neurons, both mouse (N2a) and human (NT2), co-expressed GFRα1 and c-Ret encoded receptor tyrosine kinase, whereas astrocytes expressed predominantly GFRα1, but not Ret. According to the current knowledge, the presence of Ret allows to assemble the GDNF/GFR α1/Ret complex in the lipid rafts and trigger its association with Src kinase (Airaksinen and Saarma, 2002). This association, in turn, is essential for the engagement of downstream intracellular signaling pathways, such as MAPK, PI3K or PLC-γ, known to play crucial roles in neuroprotection and neuritogenesis (Mikaels-Edman et al., 2003; Garcia-Martinez et al., 2006; Wang et al., 2007). On the other hand, in the absence of Ret, GFRα1 can associate with NCAM and activate Srclike kinase Fyn and focal adhesion kinase FAK, which are involved in the modulation of cell adhesion, migration and intercellular communication (Sariola and Saarma, 2003; Paveliev et al., 2007). The GDNF signaling requires also glucosaminoglycans, such as heparin sulphate proteoglycans, presumed to concentrate GDNF in the vicinity of the receptor complex regardless of its composition (Sariola and Saarma, 2003). Thus, neurons and astrocytes could, potentially, utilize different intracellular signaling pathways to mediate autocrine and paracrine effects of GDNF. Taken together, our results add to a growing body of evidence that engineered astrocytes might be superior biological “minipumps” to deliver neurotrophic factors, such as GDNF, into the CNS. Taking into consideration the neuron-astroglia active metabolic coupling, these cells, grafted alone or in a combination with cell replacements, should bring about the more beneficial outcomes, especially for the treatment of PD. Acknowledgment We sincerely thank Caroline Sodja for collecting Hoechst stained images of 6OHDA-treated N2a cells.

413

References Airaksinen, M.S., Saarma, M., 2002. The GDNF family: signalling, biological functions and therapeutic value. Nat. Rev. Neurosci. 3, 383–394. Akerud, P., Canals, J.M., Snyder, E.Y., Arenas, E., 2001. Neuroprotection through delivery of glial cell line-derived neurotrophic factor by neural stem cells in a mouse model of Parkinson's disease. J. Neurosci. 21, 8108–8118. Arvidsson, A., Kirik, D., Lundberg, C., Mandel, R.J., Andsberg, G., Kokaia, Z., Lindvall, O., 2003. Elevated GDNF levels following viral vector-mediated gene transfer can increase neuronal death after stroke in rats. Neurobiol. Dis. 14, 542–556. Bohn, M.C., 2004. Motoneurons crave glial cell line-derived neurotrophic factor. Exp. Neurol. 190, 263–275. Byrd, A.S., Sikorska, M., Walker, P.R., Sandhu, J.K., 2004. Effects of glutathione depletion on the viability of human NT2-derived neuronal and astroglial cultures. Neuron Glia Biology 1, 317–326. Capowski, E.E., Schneider, B.L., Ebert, A.D., Seehus, C.R., Szulc, J., Zufferey, R., Aebischer, P., Svendsen, C.N., 2007. Lentiviral vector-mediated genetic modification of human neural progenitor cells for ex vivo gene therapy. J. Neurosci. Methods. 163, 338–349. Cunningham, L.A., Su, C., 2002. Astrocyte delivery of glial cell line-derived neurotrophic factor in a mouse model of Parkinson's disease. Exp. Neurol. 174, 230–242. Ding, Y.M., Jaumotte, J.D., Signore, A.P., Zigmond, M.J., 2004. Effects of 6-hydroxydopamine on primary cultures of substantia nigra: specific damage to dopamine neurons and the impact of glial cell line-derived neurotrophic factor. J. Neurochem. 89, 776–787. Dringen, R., 2000. Metabolism and functions of glutathione in brain. Prog. Neurobiol. 62, 649–671. Dringen, R., Hirrlinger, J., 2003. Glutathione pathways in the brain. Biol. Chem. 384, 505–516. Dringen, R., Gutterer, J.M., Hirrlinger, J., 2000. Glutathione metabolism in brain metabolic interaction between astrocytes and neurons in the defense against reactive oxygen species. Eur. J. Biochem. 267, 4912–4916. Duan, D., Yang, H., Zhang, J., Zhang, J., Xu, Q., 2005. Long-term restoration of nigrostriatal system function by implanting GDNF genetically modified fibroblasts in a rat model of Parkinson's disease. Exp. Brain Res. 161, 316–324. Ericson, C., Georgievska, B., Lundberg, C., 2005. Ex vivo gene delivery of GDNF using primary astrocytes transduced with a lentiviral vector provides neuroprotection in a rat model of Parkinson's disease. Eur. J. Neurosci. 22, 2755–2764. Eslamboli, A., 2005. Assessment of GDNF in primate models of Parkinson's disease: comparison with human studies. Rev. Neurosci. 16, 303–310. Galan-Rodriguez, B., del-Marco, A., Flores, J.A., Ramiro-Fuentes, S., Gonzalez-Aparicio, R., Tunez, I., Tasset, I., Fernandez-Espejo, E., 2008. Grafts of extra-adrenal chromaffin cells as aggregates show better survival rate and regenerative effects on parkinsonian rats than dispersed cell grafts. Neurobiol. Dis. 29, 529–542. Garcia-Martinez, J.M., Perez-Navarro, E., Gavalda, N., Alberch, J., 2006. Glial cell linederived neurotrophic factor promotes the arborization of cultured striatal neurons through the p42/p44 mitogen-activated protein kinase pathway. J. Neurosci. Res. 83, 68–79. Georgievska, B., Kirik, D., Bjorklund, A., 2002. Aberrant sprouting and downregulation of tyrosine hydroxylase in lesioned nigrostriatal dopamine neurons induced by longlasting overexpression of glial cell line derived neurotrophic factor in the striatum by lentiviral gene transfer. Exp. Neurol. 177, 461–474. Gill, S.S., Patel, N.K., Hotton, G.R., O'Sullivan, K., McCarter, R., Bunnage, M., Brooks, D.J., Svendsen, C.N., Heywood, P., 2003. Direct brain infusion of glial cell line-derived neurotrophic factor in Parkinson disease. Nat. Med. 9, 589–595. Horita, Y., Honmou, O., Harada, K., Houkin, K., Hamada, H., Kocsis, J.D., 2006. Intravenous administration of glial cell line-derived neurotrophic factor gene-modified human mesenchymal stem cells protects against injury in a cerebral ischemia model in the adult rat. J. Neurosci. Res. 84, 1495–1504. Iwata-Ichikawa, E., Kondo, Y., Miyazaki, I., Asanuma, M., Ogawa, N., 1999. Glial cells protect neurons against oxidative stress via transcriptional up-regulation of the glutathione synthesis. J Neurochem 72, 2334–2344. Jakobsson, J., Lundberg, C., 2006. Lentiviral vectors for use in the central nervous system. Mol. Ther. 13, 484–493. Kafri, T., 2004. Gene delivery by lentivirus vectors an overview. Methods Mol. Biol. 246, 367–390. Karakaya, S., Kipp, M., Beyer, C., 2007. Oestrogen regulates the expression and function of dopamine transporters in astrocytes of the nigrostriatal system. J. Neuroendocrinol. 19, 682–690. Klein, S.M., Behrstock, S., McHugh, J., Hoffmann, K., Wallace, K., Suzuki, M., Aebischer, P., Svendsen, C.N., 2005. GDNF delivery using human neural progenitor cells in a rat model of ALS. Hum. Gene Ther. 16, 509–521. Kobayashi, T., Ahlenius, H., Thored, P., Kobayashi, R., Kokaia, Z., Lindvall, O., 2006. Intracerebral infusion of glial cell line-derived neurotrophic factor promotes striatal neurogenesis after stroke in adult rats. Stroke 37, 2361–2367. Kordower, J.H., Palfi, S., Chen, E.Y., Ma, S.Y., Sendera, T., Cochran, E.J., Cochran, E.J., Mufson, E.J., Penn, R., Goetz, C.G., Comella, C.D., 1999. Clinicopathological findings following intraventricular glial-derived neurotrophic factor treatment in a patient with Parkinson's disease. Ann. Neurol. 46, 419–424. Kordower, J.H., Emborg, M.E., Bloch, J., Ma, S.Y., Chu, Y., Leventhal, L., McBride, J., Chen, E.Y., Palfi, S., Roitberg, B.Z., Brown, W.D., Holden, J.E., Pyzalski, R., Taylor, M.D., Carvey, P., Ling, Z., Trono, D., Hantraye, P., Deglon, N., Aebischer, P., 2000. Neurodegeneration prevented by lentiviral vector delivery of GDNF in primate models of Parkinson's disease. Science 290, 767–773. Lin, L.F., Doherty, D.H., Lile, J.D., Bektesh, S., Collins, F., 1993. GDNF: a glial cell line-derived neurotrophic factor for midbrain dopaminergic neurons. Science 260, 1130–1132. Matsushita, N., Fujita, Y., Tanaka, M., Nagatsu, T., Kiuchi, K., 1997. Cloning and structural organization of the gene encoding the mouse glial cell line-derived neurotrophic factor, GDNF. Gene 203, 149–157.

414


McBride, J.L., During, M.J., Wuu, J., Chen, E.Y., Leurgans, S.E., Kordower, J.H., 2003. Structural and functional neuroprotection in a rat model of Huntington's disease by viral gene transfer of GDNF. Exp. Neurol. 181, 213–223. McBride, J.L., Ramaswamy, S., Gasmi, M., Bartus, R.T., Herzog, C.D., Brandon, E.P., Zhou, L., Pitzer, M.R., Berry-Kravis, E.M., Kordower, J.H., 2006. Viral delivery of glial cell linederived neurotrophic factor improves behavior and protects striatal neurons in a mouse model of Huntington's disease. Proc. Natl. Acad. Sci. U. S. A. 103, 9345–9350. Meister, A., 1988. Glutathione metabolism and its selective modification. J. Biol. Chem. 263, 17205–17208. Mikaels-Edman, A., Baudet, C., Ernfors, P., 2003. Soluble and bound forms of GFRalpha1 elicit different GDNF-independent neurite growth responses in primary sensory neurons. Dev. Dyn. 227, 27–34. Oppenheim, R.W., Houenou, L.J., Johnson, J.E., Lin, L.F., Li, L., Lo, A.C., Newsome, A.L., Prevette, D.M., Wang, S., 1995. Developing motor neurons rescued from programmed and axotomy-induced cell death by GDNF. Nature 373, 344–346. Parsadanian, A., Pan, Y., Li, W., Myckatyn, T.M., Brakefield, D., 2006. Astrocyte-derived transgene GDNF promotes complete and long-term survival of adult facial motoneurons following avulsion and differentially regulates the expression of transcription factors of AP-1 and ATF/CREB families. Exp. Neurol. 200, 26–37. Paveliev, M., Lume, M., Velthut, A., Phillips, M., Arumae, U., Saarma, M., 2007. Neurotrophic factors switch between two signaling pathways that trigger axonal growth. J. Cell Sci. 120, 2507–2516. Pertusa, M., Garcia-Matas, S., Mammeri, H., Adell, A., Rodrigo, T., Mallet, J., Cristofol, R., Sarkis, C., Sanfeliu, C., 2007. Expression of GDNF transgene in astrocytes improves cognitive deficits in aged rats. Neurobiol. Aging 29, 1366–1379. Peuchen, S., Bolanos, J.P., Heales, S.J., Almeida, A., Duchen, M.R., Clark, J.B., 1997. Interrelationships between astrocyte function, oxidative stress and antioxidant status within the central nervous system. Prog. Neurobiol. 52, 261–281. Poteryaev, D., Titievsky, A., Sun, Y.F., Thomas-Crusells, J., Lindahl, M., Billaud, M., Arumae, U., Saarma, M., 1999. GDNF triggers a novel ret-independent Src kinase familycoupled signaling via a GPI-linked GDNF receptor alpha1. FEBS Lett. 463, 63–66. Raicevic, N., Mladenovic, A., Perovic, M., Miljkovic, D., Trajkovic, V., 2005. The mechanisms of 6-hydroxydopamine-induced astrocyte death. Ann. N. Y. Acad. Sci. 1048, 400–405. Rosenblad, C., Georgievska, B., Kirik, D., 2003. Long-term striatal overexpression of GDNF selectively downregulates tyrosine hydroxylase in the intact nigrostriatal dopamine system. Eur. J Neurosci. 17, 260–270. Saarma, M., Sariola, H., 1999. Other neurotrophic factors: glial cell line-derived neurotrophic factor (GDNF). Microsc. Res. Tech. 45, 292–302.

Sandhu, J.K., Sikorska, M., Walker, P.R., 2002. Characterization of astrocytes derived from human NTera-2/D1 embryonal carcinoma cells. J. Neurosci. Res. 68, 604–614. Sandhu, J.K., Pandey, S., Ribecco, M., Monette, R., Borowy-Borowski, H., Walker, P.R., Sikorska, M., 2003. Molecular mechanisms of glutamate neurotoxicity in mixed cultures of NT2-derived neurons and astrocytes: protective effects of coenzyme Q10. J. Neurosci. Res. 72, 691–703. Sariola, H., Saarma, M., 2003. Novel functions and signalling pathways for GDNF. J. Cell Sci. 116, 3855–3862. Sherer, T.B., Fiske, B.K., Svendsen, C.N., Lang, A.E., Langston, J.W., 2006. Crossroads in GDNF therapy for Parkinson's disease. Mov. Disord. 21, 136–141. Takuma, K., Baba, A., Matsuda, T., 2004. Astrocyte apoptosis: implications for neuroprotection. Prog. Neurobiol. 72, 111–127. Trendelenburg, G., Dirnagl, U., 2005. Neuroprotective role of astrocytes in cerebral ischemia: focus on ischemic preconditioning. Glia 50, 307–320. Trono, D., 2000. Lentiviral vectors: turning a deadly foe into a therapeutic agent. Gene Ther. 7, 20–23. Trupp, M., Scott, R., Whittemore, S.R., Ibanez, C.F., 1999. Ret-dependent and independent mechanisms of glial cell line-derived neurotrophic factor signaling in neuronal cells. J. Biol. Chem. 274, 20885–20894. Wang, H.J., Cao, J.P., Yu, J.K., Gao, D.S., 2007. Role of PI3-K/Akt pathway and its effect on glial cell line-derived neurotrophic factor in midbrain dopamine cells. Acta Pharmacol. Sin. 28, 166–172. Wang, L.J., Lu, Y.Y., Muramatsu, S., Ikeguchi, K., Fujimoto, K., Okada, T., Mizukami, H., Matsushita, T., Hanazono, Y., Kume, A., Nagatsu, T., Ozawa, K., Nakano, I., 2002. Neuroprotective effects of glial cell line-derived neurotrophic factor mediated by an adeno-associated virus vector in a transgenic animal model of amyotrophic lateral sclerosis. J. Neurosci. 22, 6920–6928. Wilson, J.X., 1997. Antioxidant defense of the brain: a role for astrocytes. Can. J. Physiol. Pharmacol. 75, 1149–1163. Wong, L.F., Goodhead, L., Prat, C., Mitrophanous, K.A., Kingsman, S.M., Mazarakis, N.D., 2006. Lentivirus-mediated gene transfer to the central nervous system: therapeutic and research applications. Hum. Gene Ther. 17, 1–9. Zhao, Z., Alam, S., Oppenheim, R.W., Prevette, D.M., Evenson, A., Parsadanian, A., 2004. Overexpression of glial cell line-derived neurotrophic factor in the CNS rescues motoneurons from programmed cell death and promotes their long-term survival following axotomy. Exp. Neurol. 190, 356–372. Zurn, A.D., Widmer, H.R., Aebischer, P., 2001. Sustained delivery of GDNF: towards a treatment for Parkinson's disease. Brain Res. Brain Res. Rev 36, 222–229.