Fetal striatum- and ventral mesencephalon–derived expanded neurospheres rescue dopaminergic neurons in vitro and the nigro-striatal system in vivo

Neuroscience 154 (2008) 606 – 620

FETAL STRIATUM- AND VENTRAL MESENCEPHALON– DERIVED EXPANDED NEUROSPHERES RESCUE DOPAMINERGIC NEURONS IN VITRO AND THE NIGRO-STRIATAL SYSTEM IN VIVO D. MOSES,a,d J. DRAGO,a,b* Y. TEPER,a I. GANTOIS,a D. I. FINKELSTEINa,c AND M. K. HORNEa

cue of nigral dopaminergic neurons. Identification of NS-derived soluble factor(s) may lead to development of novel neuroprotective therapies for PD. An unexpected observation of the present study was the detection of the ectopic host-derived tyrosine hydroxylase (TH) – expressing cells in sham-grafted mice and NSE11- and NSvm -grafted mice. We speculate that injury-derived signals (such as inflammatory cytokines that are commonly released during transplantation) induce TH expression in susceptible cells. Crown Copyright © 2008 Published by Elsevier Ltd on behalf of IBRO. All rights reserved.

a

Howard Florey Institute, University of Melbourne, Parkville, Victoria, 3010, Australia

b

Centre for Neuroscience, the University of Melbourne, Parkville, Victoria, 3010, Australia

c

Mental Health Research Institute of Victoria, 155 Oak Street, Parkville, Victoria, 3052, Australia

d

Centre for Clinical Neurosciences and Neurological Research, St. Vincent’s Hospital, Fitzroy, Victoria 3065, Australia

Key words: neurospheres, transplantation, substantia nigra, GFP, dopaminergic, rescue. Abstract—The pathogenesis of Parkinson’s disease (PD) involves ongoing apoptotic loss of dopaminergic neurons in the substantia nigra pars compacta. Local delivery of the trophic factors can rescue dopaminergic neurons and halt the progression of PD. In this study we show that fetal E11 striatum-derived neurospheres and E14.5 ventral mesencephalon (VM) – derived neurospheres (NSE11 and NSvm, respectively) are a source of factors that rescue dopaminergic neurons. First, long-term expanded NSE11 and NSvm rescued primary dopaminergic neurons from serum-deprivation induced apoptosis and promoted survival of dopaminergic neurons for 14 days in vitro and this effect was due to soluble contact-independent factor/s. Second, green fluorescent protein– expressing NSE11 and NSvm grafted into the midbrain of mice with unilateral 6-hydroxydopamine-induced Parkinsonism resulted in partial rescue of the nigro-striatal system and improvement of the hypo-dopaminergic behavioral deficit. Reverse transcription-polymerase chain reaction (RT-PCR) analysis demonstrated that intact NSE11 and NSvm expressed fibroblast growth factor-2, brain-derived neurotrophic factor (BDNF), pleiotrophin, neurotrophin-3, but not glial cell line– derived neurotrophic factor (GDNF). GDNF expression was also undetectable in vivo in grafted NSE11 and NSvm suggesting that NS-derived factor/s other than GDNF mediated the res-

Parkinson’s disease (PD) is a neurodegenerative condition characterized by progressive apoptosis of dopaminergic (DA) neurons in the substantia nigra pars compacta (SNpc) resulting in DA denervation of the striatum, the target structure for SNpc DA neurons (Hornykiewicz, 1993). Ongoing loss of nigral DA neurons is believed to underpin the progression of PD symptoms after the initial 5–7 years of the disease (Hughes et al., 1993; Lang and Lozano, 1998a,b; Hornykiewicz, 2001). PD patients typically display a loss of 50 – 60% of nigral DA neurons at the time of initial presentation (Hornykiewicz, 1993). Rescuing the remaining 40 –50% of nigral DA neurons may be sufficient to halt clinical disease progression and maintain patients with a mild degree of disability and a sustained response to levodopa. Thus far, promising results have been obtained with glial cell line-derived neurotrophic factor (GDNF) (Barker, 2006). Despite this finding, there is an ongoing demand for molecules with neuroprotective effect on DA neurons (Rascol et al., 2002). In this context, committed neural precursors may prove useful. While grafting of neural precursor cells (NPCs) has been extensively investigated in the context of cell-replacement therapy, recent studies have shown that undifferentiated NPCs may act via the so-called “bystander” effect (Ourednik et al., 2002; Martino and Pluchino, 2006), which refers to a therapeutic benefit not due to NPCderived cell replacement per se, but due to NPC-derived neuromodulatory and neurotrophic signals that mitigate pivotal elements of the pathological process (Ader et al., 2001; Park et al., 2002; Pluchino et al., 2003; Martino and Pluchino, 2006). Previous in vivo studies identified the “bystander” rescue effect of immortalized neural stem cell (NSC) lines on the host DA system in 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) –induced Parkinsonism (Ourednik et al., 2002). A recently published study has demonstrated

*Correspondence to: J. Drago, Howard Florey Institute, University of Melbourne, Parkville, Victoria, 3010, Australia. Tel: ⫹613-8344-1959; fax: ⫹613-9347-0446. E-mail address: [email protected] (J. Drago). Abbreviations: ANOVA, analysis of variance; BDNF, brain-derived neurotrophic factor; BrdU, 5-bromo-2-deoxyuridine; CDNF, conserved dopamine neurotrophic factor; CMvm, medium conditioned by NSvm; CME11, medium conditioned by NSE11; CPu, caudate-putamen; DA, dopaminergic; DAT, dopamine transporter; EGF, epidermal growth factor; FGD, fluorogold; FGF-2, fibroblast growth factor -2; GDNF, glial cell line– derived neurotrophic factor; GFP, green fluorescent protein; MANF, mesencephalic astrocyte-derived neurotrophic factor; MPTP, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine; NF, neurofilament; NPC, neural precursor cells; NS, neurosphere; NSC, neural stem cell; NSvm, ventral mesencephalon neurospheres derived at E14.5; NSE11, striatal neurospheres derived at E11; NT-3, neurotrophin-3; PBS, phosphatebuffered saline; PD, Parkinson’s disease; PFA, paraformaldehyde; PLL, poly-L-lysine; PTN, pleiotrophin; RS, rotational score; RT-PCR, reverse transcription–polymerase chain reaction; SNpc, substantia nigra pars compacta; SVZ, subventricular zone; TH, tyrosine hydroxylase; VM, ventral mesencephalon; 6-OHDA, 6-hydroxydopamine.

0306-4522/08$32.00⫹0.00 Crown Copyright © 2008 Published by Elsevier Ltd on behalf of IBRO. All rights reserved. doi:10.1016/j.neuroscience.2008.03.058

606

D. Moses et al. / Neuroscience 154 (2008) 606 – 620

that simultaneous grafting of the genetically unmodified human NPC into the caudate-putamen (CPu) and midbrain improves behavioral impairment in primates with MPTPinduced Parkinsonism mainly due to “bystander” mechanism (Redmond et al., 2007). GDNF appears to be a central player that mediates the rescue effect of undifferentiated NPC on host DA neurons (Ourednik et al., 2002; Redmond et al., 2007). Fetal- and adult-derived NPC can be maintained, propagated and expanded without genetic manipulations ex vivo using a serum-free neurosphere (NS) culture system (Reynolds and Weiss, 1992). We aimed to explore if genetically unaltered NS are a source of factors that rescue nigral DA neurons. Previous in vitro studies revealed enhanced survival and functional capacity of DA neurons in co-culture with expanded NS (Ostenfeld et al., 1999) and this effect could be mediated by soluble factors other than GDNF (Rafuse et al., 2005; Moses et al., 2006). Intra-striatal grafting of subventricular zone (SVZ) – derived NS progenitors alleviated behavioral parkinsonian deficits (Richardson et al., 2005); however SVZ-derived expanded NS were teratogenic in vivo (Rafuse et al., 2005) in contrast to fetal-derived NS that usually lack in vivo teratogenic potential (Uchida et al., 2000; Reynolds and Rietze, 2005). In this study we examined the capacity of fetal-derived NS to rescue mesencephalic DA neurons under conditions of serum deprivation–induced apoptosis. For that purpose we derived and expanded NS from the developing ventral mesencephalon (VM), a brain region that contains DA neuron cell bodies, and the developing striatum, the innervation target of DA neurons. We further explored the efficacy of NS-derived factors in an in vivo “rescue” paradigm using the 6-hydroxydopamine (6-OHDA) rodent model of hemi-Parkinsonism and delayed engraftment of NS into the area of lesioned SN. The “rescue” paradigm is clinically relevant since it resembles the symptomatic stage of PD, characterized by established degeneration of DA neurons (Bowenkamp et al., 1996; Bjorklund et al., 1997; Kirik et al., 2001; Wang et al., 2002). Intra-striatal grafting may not be the optimal method for that purpose as 6-OHDA lesions cause retraction of nigrostriatal projections (Rosenblad et al., 2000) essentially preventing the retrograde axonal transport of intra-striatally delivered trophic factors to the cell bodies of nigral DA neurons (Tomac et al., 1995; Rosenblad et al., 1998; Kirik et al., 2001), whereas grafting into the lesioned midbrain may facilitate access of NSderived factors to nigral DA neurons. Earlier studies investigated grafting of single cell suspension into the midbrain, but were met with unsatisfactory graft survival (Bjorklund et al., 1983; Dunnett and Björklund, 1992; Freeman et al., 1995). In the present study, NPCs were engrafted as intact NS because we anticipated that this approach could confer an in vivo survival advantage to grafted NS as the complex three-dimensional organization of intact NS comprising biologically active factors, such as extracellular matrix, growth factors and cytokines (Benoit et al., 2001; Campos, 2004; Deleyrolle et al., 2006; Martino and Pluchino, 2006) would not be disrupted.

607

We report that expanded fetal-derived NS produce soluble factors that support long-term survival of DA neurons in vitro. In vivo, NS-derived activity increased the number of surviving DA neurons in the SN resulting in DA re-innervation of the ipsilateral striatum and partial restoration of the hypo-DA behavioral deficit. Grafts of NSvm did not differentiate into DA neurons, confirming that the rescue effect was on the host DA system rather than mediated by graft derived cell-replacement.

EXPERIMENTAL PROCEDURES All methods conformed to the Australian National Health and Medical Research Council published code of practice for the use of animals in research and approved by the Howard Florey Institute animal ethics committee. All experiments conformed to international guidelines on the ethical use of animals. All efforts were made to minimize the number of animals used and their suffering.

Generation and long-term expansion of NS NS were derived from E11 or E14.5 embryos of time-pregnant C57BL/6J mice. Matings are set up and mice examined for the presence of the vaginal plug at 08:00 h, the date the vaginal plug is first identified is designated as E0. Alternatively, fetal NS were derived from transgenic mice expressing enhanced green fluorescent protein (GFP⫹) under the control of a chicken beta-actin promoter and cytomegalovirus enhancer (Jackson Laboratories, Bar Harbor, ME, USA) (Okabe et al., 1997). It is important to note that GFP⫹ mice were also on a C57BL/6J background. Previous studies demonstrated NSCs derived from this GFP mouse strain maintain GFP expression upon terminal differentiation (Mizumoto et al., 2003; Jensen et al., 2004). Time-pregnant mice were killed by cervical dislocation and embryonic sacs were dissected and collected in ice-cold hibernation medium (phosphate-buffered saline (PBS) supplemented with Hepes 10 mM (Gibco, Invitrogen, Carlsbad, CA, USA), sodium pyruvate 1 mM (Sigma Aldrich Corp., St. Louis, MO, USA), 0.6% D-glucose and penicillin/streptomycin (Gibco, 50 IU/50-␮g per ml). For derivation of E11 striatal NS, whole brain was extracted and forebrain was separated from the mesencephalon followed by isolation of the ventral forebrain using a dissection microscope. For generation of NS derived from VM, the mesencephalic flexure of the E14.5 embryonic brain was identified and dissected, and the VM was separated from the dorsal mesencephalon along the line of the presumptive sulcus limitans. Harvested tissue was pooled in hibernation medium and enzymatically dissociated using serum-free conditions (Drago et al., 1991; Moses et al., 2006). Initially, 5⫻104 cells/ml were plated in uncoated 25 cm2 flasks (BD Falcon, Heidelberg, Germany) at a final volume of 5 ml of NS medium and incubated at 37 °C in humidified 5% CO2/95% atmosphere air incubator. The NS medium was composed of DMEM/F12/2.5 mM glutamine (Gibco, cat. No.: 11320 – 033), 1% N2 supplement (Gibco), 0.3% D-glucose, penicillin/streptomycin, 10 ng/ml of epidermal growth factor (EGF, Peprotech) and 10 ng/ml of fibroblast growth factor-2 (FGF-2, Peprotech, Rocky Hill, NJ, USA). Cultures were supplemented with 20% of the fresh NS medium and 100% FGF-2/EGF every 48 h. Under these conditions primary NS were formed within 7–9 days. Primary and all subsequent NS were propagated by enzymatic dissociation with trypsin and seeded at a density 5⫻104 cells/ml in the presence of FGF-2/EGF resulting in formation of the subsequent generation of NS within 6 –7 days. Intact NS were plated onto poly-L-lysine (PLL, Sigma Aldrich Corp.) – coated chamber-slides for assessment of tripotentiality as previously described (Moses et al., 2006).

608

D. Moses et al. / Neuroscience 154 (2008) 606 – 620

Preparation of primary cultures of mesencephalic DA neurons and co-culture with NS and conditioned medium The VM was dissected from E14.5 embryos and a single cell suspension prepared using enzymatic dissociation as described above. Serum- and mitogen-free conditions were used throughout this experiment. Primary VM cultures were co-cultured side-byside with striatal neurospheres derived at E11 (NSE11) or ventral mesencephalon neurospheres derived at E14.5 (NSvm). Primary VM cells and NS were plated in 4-well PLL-coated chamber-slides (BD Biosciences) using a micro-islet technique: suspension of primary VM cells (1.2⫻105 cells in the final volume of 25 ␮l of serum- and mitogen-free NS medium) was carefully placed at the pole of the well using a yellow pipette tip; 10 –15 intact wild-type or GFP⫹ NS were collected in the final volume of 50 ␮l of NS medium under the guidance of a microscope and plated at the opposite pole of the same well using a blue pipette tip; special care was taken to avoid mixing NS with primary DA neurons at the time of plating. Cells in micro-islets were allowed to attach for 3– 4 h; cultures were then washed twice with mitogen-free NS medium and incubated in serum- and mitogen-free NS medium. The medium was replaced at the 7-day time-point. For preparation of conditioned medium (CM), approximately 200 –250 expanded NSE11 or NSvm were collected, washed twice with PBS and re-suspended in 5 ml of serum- and mitogen-free NS medium for 48 –72 h. Supernatant was collected and after centrifugation at 7000⫻g applied fresh to primary VM cultures. For detection of the putative proliferating DA neuroblasts, 5-bromo-2-deoxyuridine (BrdU, Sigma Aldrich Corp) was applied to co-cultures at the range of concentrations 1–10 ␮M as indicated in the “Results” section. To ensure reproducibility, all experiments were repeated using expanded NS that were derived from at least eight different litters.

RNA preparation and reverse transcription–polymerase chain reaction (RT-PCR) RNA was extracted from expanded NS. Gene expression was characterized on pooled NS (10 –15). Each PCR reaction was run three separate times. RNA isolated from embryonic brain was used as a positive control. RNA was isolated using Trizol (Life Technologies-BRL) and cDNA synthesis was undertaken using the superscript first-strand synthesis system (Life Technologies-Invitrogen). All cDNA amplifications were undertaken in a final volume of 50 ␮l containing 2.0 mM MgCl2, 0.84 ␮M of each custom primer (Geneworks, Adelaide, Australia), 0.5 mM dNTP (Promega, Madison, WI, USA) mixture and two units of Taq Polymerase in PCR Buffer (Promega) using the T3-Thermocycler (Biometra, Göttingen, Germany). The PCR protocol consisted of an initial 5 min denaturation step at 95 °C followed by 35 cycles of: 1 min annealing step at 58 °C, a 1 min

extension step at 70 °C followed by a 45 s 95 °C denaturation step. The run was concluded by a final 10-min 70 °C extension. The primer sequences used for detection of GDNF, brain-derived neurotrophic factor (BDNF), pleiotrophin (PTN), neurotrophin-3 (NT-3) and FGF-2 are shown in Table 1.

Immunocytochemistry and in vitro cell counting procedure Immunocytochemistry and in vitro cells counts were obtained as previously described (Moses et al., 2006). Briefly, differentiated NS cultures or primary VM co-cultures were fixed for 15 min with 4% paraformaldehyde (PFA) in 0.1 M PBS, pH 7.3. Fixed cells were permeabilized by incubation in 100% methanol at ⫺20 °C for 10 min. For GalC and NG2 staining, cultures were fixed for 8 min with 4% PFA and the methanol step was omitted. Blocking solution consisting of 10% normal goat serum in PBS (NGS/PBS) was applied for 1 h prior to incubation with primary antibody. Primary antibodies were: neuron-specific (Draberova et al., 1998) mouse monoclonal anti-␤-tubulin-III (clone TU-20, MAB1637, Chemicon), 1:400; rabbit polyclonal anti-Tuj1 (Covance), 1:800; mouse monoclonal anti-neurofilament-160-200 kDa (NF) antibody (Zymed), 1:400; rabbit polyclonal anti-GFAP (Chemicon), 1:1000; mouse monoclonal anti-GalC (Chemicon), 1:200; mouse polyclonal anti– tyrosine hydroxylase (TH) (Chemicon), 1:1000; mouse monoclonal anti-nestin (Chemicon), 1:200; rabbit polyclonal anti-NG2 (Chemicon), 1:200; mouse monoclonal anti-BrdU (Sigma), 1:50; mouse monoclonal fluorescein-conjugated anti-BrdU (Abcam), 1:10. Mouse monoclonal radial glia marker anti-RC2 antibody (IgM, Developmental Studies Hybridoma Bank) was applied at 1:200. Cultures were blocked again prior to application of appropriate secondary antibodies. Hoescht 33258 nuclear stain (5-␮g/ ml) was applied to cultures for 10 min. Individual differentiated NS colonies or primary VM cultures were identified and delineated by Hoescht⫹ nuclei. The estimate of the Hoescht⫹ cell nuclei and the proportion of cells of specific phenotype were estimated using a random sampling procedure (Stereo Investigator Program, MicroBrightField, VT, USA).

6-OHDA administration, NS grafting and quantification of rotational asymmetry Young adult (10 –12 week) C57BL/6J mice were anesthetized with 4% chloral hydrate in PBS (10 ml/kg, i.p.); a single stereotaxic injection of DA neurotoxin 6-OHDA (2 ␮l of 5 ␮g/␮l solution in 0.9% normal saline/0.02% ascorbic acid) was made into the area of right SNpc (anteroposterior, 3.1 mm; lateral, 1.2 mm; dorsoventral, 4.7 mm, with respect to bregma) (Franklin and Paxinos, 1997). Mice were pretreated with desipramine hydrochloride (25 mg/kg, Sigma) by i.p. injection 30 min prior to 6-OHDA injection to reduce damage of noradrenergic neurons. After surgery, the skin was sutured, antiseptic (1% w/w iodine, Betadine) was

Table 1. Primer sequences Gene

Primer sequence, forward (f)/backward (b)

Amplicon size

Accession No./Ref

BDNF

f. ATGGGACTCTGGAGAGCGTGAA b. CGCCAGCCAATTCTCTTTTTGC f. CGCTGACCAGTGACTCCAATATGC b. GTTAGCCTTCTACTCCGAGACAGG f. GACACTGCTGAGGCCGGGAA b. TGCCCTTTTCCTGGTCCACA f. AGAGCGACCCACACGTCAAAC b. CCAACTGGAGTATTTCCGTGACC f. CTTATCTCCGTGGCATCCAA b. TCTGAAGTCAGTGCTCGGACGT

600 bp

Promega kit

350 bp

O’Rourke et al., 1999

400 bp

NM_008973

200 bp

Jin et al., 2002

370 bp

NM_008742

GDNF PTN FGF-2 NT-3

D. Moses et al. / Neuroscience 154 (2008) 606 – 620 applied to the wound, and the mice left in a warmed cage to recover. Paracetamol (100 mg/kg) was administered routinely in drinking water as an analgesic after surgery. Rotational score (RS) in response to i.p. administration of 5 mg/kg of d-amphetamine (Sigma) was determined at 3 weeks after 6-OHDA injection, and grafting of NS was carried out next day. Mice were allowed to habituate for 30 min prior to administration of d-amphetamine; rotations were video-recorded and analyzed. RS was defined as the net number (right minus left) of complete 360° turns over a 90 min period. Approximately 8 to 10 NS with approximate diameters of 200 –250 ␮m were individually drawn into a sterile glass cannula (external diameter 1 mm, internal diameter 0.7 mm) and grafted into the right mesencephalon. Transplantation cannula was first lowered to the same dorso-ventral coordinate as for 6-OHDA injection and then slowly withdrawn by 0.7 mm with the intention of depositing graft as close to SN as possible while avoiding excessive injury to SN. Control mice received phosphate buffered saline (PBS) through the same glass cannula. Rotational asymmetry was determined at 4, 8, 12 and 22 weeks after grafting. In preliminary experiments we determined an approximation of the number of cells in intact NS. For that purpose intact NS were fixed, embedded into OCT compound and cryo-sectioned at 14 ␮m thickness to determine cell density according to nuclear Hoescht 33258 stain (Fig. 1C). Based on these calculations, we estimated that one NS of 200 ␮m diameter contained between 9000 –15,000 cells.

609

dures were not applicable (West and Gundersen, 1990; West et al., 1991). Instead we determined the absolute number of TH⫹ cells in the lesioned SN quadrant defined anatomically as the region bound by the third cranial nerve rootlet, the ventral border of the mesencephalon and the dorsal border of SNpc. To delineate this region, overlays were prepared from the normal brain of age-matched animals with the assistance of the Stereo Investigator Program (MicroBrightField) and superimposed on corresponding sections using prominent landmarks such as ventricles and hippocampus. Accordingly, counting was restricted to sections containing the third cranial nerve rootlet. Five to six sections were counted per mouse in all cases except for two mice grafted with E11-derived NS in which four sections were obtained. The number obtained was then multiplied by three to account for the entire antero-posterior extent of SNpc. Counting of TH⫹ cells in intact SN was undertaken using the same method. The calculated total counts of TH⫹ neurons in SN were influenced by the counting method. Using the same method, the number of TH⫹ neurons in non-lesioned age-matched mice was approximately 25% of the number ordinarily obtained with formal stereological counting methods (Parish et al., 2001, 2005). Estimates of the density of striatal DAT expressing (DAT⫹) terminals were obtained according to previously published methods (Parish et al., 2001). In brief, stereological estimates of DAT⫹ varicosities in the dorsal CPu were made at regular intervals (x⫽170 ␮m, y⫽170 ␮m) using a counting frame (5⫻4 ␮m⫽20 ␮m2) and values of ⬍0.1 were accepted for coefficients of error (CE) and coefficients of variance (CV) (West and Gundersen, 1990). The density of DAT⫹ terminals in the striatum is an established measure of DA innervation (Ciliax et al., 1995; Freed et al., 1995).

Immunohistochemistry Mice were killed by overdose of sodium pentobarbitone and transcardially perfused as described (Parish et al., 2001). Brains were removed and equilibrated in 30% sucrose solution in PBS at 4 °C for 24 – 48 h. A one in 15 series of 16-␮m coronal sections was obtained through the striatum. A one in three series of coronal 25-␮m sections was obtained through the SNpc. Sections were mounted directly onto “Superfrost” slides. The immunolabeling procedure was conducted as described in detail elsewhere (Finkelstein et al., 2000). Rabbit polyclonal anti-TH antibody (Chemicon) was applied at 1:1000 to identify midbrain DA neurons; anti-dopamine transporter (DAT) antibody (Chemicon) was used at a 1:3000 dilution to identify striatal DA terminals. Briefly, sections were incubated with primary antibody for 24 – 48 h followed by biotinylated secondary antibody, avidin– biotin–peroxidase complex and visualization with cobalt and nickel-intensified diaminobenzidine reaction. Sections were counterstained with Neutral Red. For immunofluorescent double-labeling, sections were incubated in blocking solution (10% goat serum/PBS/0.3% Triton X-100) for 1 h at room temperature and then incubated in primary antibody diluted in blocking solution at 4 °C for 24 h. The sections were then rinsed in PBS 10 min ⫻3 and blocked again for 30 min followed by application of secondary goat anti-rabbit 594 (ab=) 2-fragment antibody (Molecular Probes;1:200) for 1 h at room temperature. Goat anti-GFP antibody directly conjugated to 488fluorophore (Molecular Probes; 1:100) was included with secondary 594 antibody. Primary polyclonal antibodies were rabbit anti-TH (Chemicon; 1:800), rabbit anti-GFAP (Dako; 1:800), rabbit anti-NG2 (Chemicon; 1:200), and rabbit anti-GDNF (Serotec 1:200). Ki67 monoclonal antibody (Dako) was applied in the range of dilutions 1:50 –1:200.

Quantification of THⴙ neurons in SN and DATⴙ terminals in dorsal CPu Because of the low number of counting objects (i.e. the number of TH⫹ cells in the lesioned SNpc), standard stereological proce-

Calculations and statistics Data were expressed as mean⫾S.E.M. Data analysis was undertaken with SPSS 11.0 for Windows or GraphPad Prism 4 statistical software. Changes of RS over time as a function of grafting with NSE11, NSvm and controls were compared with two-way analysis of variance (ANOVA) for repeated measures using general linear models statistical procedure. Significance of differences on each test day was determined by pairwise comparisons test with Bonferroni correction for multiple comparisons. Significance of difference between other mean values was determined with one-way ANOVA with Dunnet’s post hoc test. Probability (P) values greater than 5% were considered non-significant.

RESULTS Characterization of the long-term expanded NS derived from E11 striatum and E14.5 VM NS culture system allows derivation of NPC at selected developmental stages (Ostenfeld et al., 2002) from almost any region of the CNS. Long-term in vitro expansion of primary NS is necessary in order to obtain a sufficient quantity of NS for transplantation and isolation of trophic factors. However, the gene expression profile of the expanded NS is altered compared with the primary NS (Santa-Olalla et al., 2003); this is also accompanied by the shift of cellular differentiation pattern toward a more gliogenic potential. Thus, it is important to examine whether these characteristics continue to vary from passage to passage in established long-term expanded NS cultures. To that end, primary NSvm were derived and propagated in the presence of FGF-2 and EGF for up to 6 months. Intact NSvm were differentiated by mitogen withdrawal following plating on PLL-coated surface (see Experimental Proce-

610

D. Moses et al. / Neuroscience 154 (2008) 606 – 620

Fig. 1. In vitro characterization of NSE11 and NSvm. (A) Intact GFP⫹ NS visualized with light microscopy and (B) same NS under epifluorescence (scale bars⫽50 ␮m). NS were also embedded in OCT compound and sectioned. (C) Cell nuclei in sectioned NS were identified with Hoechst 33258 (5-␮g/ml); scale bar⫽10 ␮m. Secondary NSE11 differentiate into neurons. (D) Tuj1⫹ neurons (red), scale bar⫽20 ␮m. (E) Secondary NSE11 also produced TH⫹ neurons. (F) Tuj1⫹ neurons (red) and GFAP⫹ astrocytes (green) in differentiated long-term passaged NSE11, scale bar⫽25 ␮m. (G) NS-derived GalC⫹ oligodendrocytes, scale bar⫽80 ␮m. (H) Growth factor expression profile in pooled NSE11 and NSvm was examined with PCR. RNA was extracted from long-term passaged NS. As a positive control, similar RT-PCR procedure was carried out on RNA isolated from embryonic brain. NSE11 and NSvm expressed BDNF and PTN, but not GDNF. For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.

dures). We examined the differentiation pattern of primary and secondary NSvm at every fifth passage thereafter. As previously reported, only primary and secondary NSvm produced a small number of Tuj1⫹ neurons, while the vast majority of differentiated NS cells expressed markers of glial lineage (Moses et al., 2006). Starting from passage 7, NSvm differentiated only into the cells of the glial lineage and no Tuj1⫹ or NF⫹ neurons could be detected (data not shown). The pattern of cellular differentiation of the expanded NSvm was essentially similar irrespective of passage: GFAP⫹ astrocytes accounted for approximately

60% of the differentiated cells; NG2⫹ cells were the second largest cell population in differentiated NSvm constituting 15–20%; a small proportion of GalC positive oligodendrocytes and single RC2⫹ and Nestin⫹ cells were also present (data not shown). In parallel experiments we derived and expanded under identical conditions NSE11 from the embryonic E11 striatum. Secondary NSE11 differentiated into Tuj1⫹ neurons (38⫾5%), GFAP⫹ astrocytes (39⫾4%) or GalC⫹ oligodendrocytes (7⫾2%) (Fig. 1D, G) if induced to differentiate by mitogen withdrawal. Single TH⫹ cells

D. Moses et al. / Neuroscience 154 (2008) 606 – 620

were identified in some differentiated primary and secondary NSE11 (Fig. 1E). As expected, the neuronogenic potential of NSE11 substantially declined with long-term expansion. The differentiation pattern of NSE11 became stable starting from passage 12: expanded NSE11 consistently produced small proportion of neurons (two to four Tuj1⫹ neurons/NS), while GFAP⫹ astrocytes constituted the majority of differentiated cells (Fig. 1F). No TH⫹ cells were found in long-term expanded NSE11. The expression profile of the selected growth factors and cytokines was examined in long-term expanded (⬎16 passages) NSvm and NSE11 by RT-PCR. Both NSE11 and NSvm expressed BDNF, PTN, FGF-2 and NT-3 (data not shown), but not GDNF (Fig. 1H), and this expression pattern persisted with further passages. Thus, long-term expanded NS derived from a given CNS region were quite stable from passage to passage with respect to cellular differentiation and gene expression. Expanded NSE11 displayed properties of NSCs, such as the capacity for long-term self-renewal and a potential for differentiation into neurons, astrocytes and oligodendrocytes, whereas expanded NSvm were incapable of neuronal differentiation. In all subsequent experiments NSvm and NSE11 expanded beyond 16 passages were used. NSvm- and NSE11-derived factors rescue primary DA neurons from apoptosis induced by serum deprivation We next examined the in vitro effect of NSvm and NSE11derived factors on mesencephalic DA neurons. DA neurons undergo rapid apoptosis under serum-free culture conditions (Takeshima et al., 1994). This culture model is widely used to study the rescue effect of a given molecule (Lin et al., 1993; Schierle et al., 1999; Goggi et al., 2000; Sawada et al., 2000; Andres et al., 2005). We co-cultured primary DA neurons with intact NSvm or NSE11 using a side-by-side co-culture paradigm. This technique allows simultaneous assessment of the role of contact-dependent and contact-independent NSvm- and NSE11-derived factors in survival of DA neurons. It also minimizes intermixing of NS-derived cells and target primary VM neurons, but permits sharing of diffusible molecules between co-cultured cell populations. We plated intact NS and primary cultures separately at the opposite poles of the same well using a micro-islet technique (see Experimental Procedures). In some experiments primary DA neurons were co-cultured with long-term expanded NSE11 and NSvm derived from GFP⫹ mice in order to identify cells of a given source. Examination of the cocultures under epifluorescence consistently demonstrated a minimal overlap zone of GFP⫹ and GFP-negative cells, while beyond this area either GFP⫹ NS cells or GFPnegative primary VM cells occupied the respective half of the well. We found that both NSE11 and NSvm promoted the survival of DA neurons for 14 days, the longest time point examined (Fig. 2A, B, H, I). As previously reported, no TH⫹ cells were detected by day 7 when cells were cul-

611

tured in NS medium alone (Takeshima et al., 1994; Moses et al., 2006). We established that NS-derived contact-dependent factors were not required for the survival of TH⫹ neurons, as TH⫹ neurons were not restricted to the overlap zone with NSvm or NSE11, but were distributed throughout the primary VM cultures. This was further confirmed using co-cultures of GFP-negative primary DA neurons with GFP⫹ NS, which revealed that the majority of TH⫹ neurons were neither in contact nor in close proximity to GFP⫹ NS. Application of medium conditioned by NSE11 (CME11) or NSvm (CMvm) to primary VM cultures also increased survival of DA neurons confirming that NS produced soluble factors that rescued DA neurons (Fig. 2D–G). The number of the surviving TH⫹ neurons was ⬃30% less in CME11 and CMvm experiments compared with co-cultures (411⫾24 and 312⫾18 of TH⫹ neurons/well, respectively by day 7 in vitro; n⫽8 –12). Interestingly, a small proportion of NSE11-derived GFP⫹ cells with neuronal morphology also expressed TH in cocultures with GFP-negative primary DA neurons (Fig. 2C); in these cases GFP⫹/TH⫹ cells were always juxtaposed to primary culture-derived cells. GFP⫹/TH⫹ neurons did not incorporate BrdU suggesting that primary VM culture-derived factors acted on NSE11-derived committed neuronal precursors to induce DA phenotype (data not shown). We further investigated whether NS-derived factors were capable of maintaining proliferation of the putative DA neuroblasts in primary VM cultures. To that end, cultures were pulsed with BrdU on day 2, day 5 and day 7 in vitro and fixed 48 h later. We detected incorporation of BrdU in Nestin⫹ and Tuj1⫹ neurons, but no TH⫹ neurons incorporated BrdU (data not shown). The number and the proportion of TH⫹ neurons in vitro decreased over time (Fig. 2H, I) suggesting that NSderived factors were incapable of inducing the DA phenotype. Primary VM cultures comprise a heterogeneous cellular population with only a small proportion of cells being DA neurons. Previous studies have suggested that non-neuronal components of primary VM cultures participate in promoting the survival of DA neurons. For example, the survival effect of serum or FGF-2 on primary DA neurons is indirect and potentially mediated by an expanded glial population in primary VM cultures (Engele and Bohn, 1991; O’Malley et al., 1992, 1994). In this study we implemented serumfree conditions that are known to prevent expansion of astrocytes in primary VM cultures (Takeshima et al., 1994). Consistent with these data we found that GFAP⫹ astrocytes constituted ⬍5% of primary VM culture-derived cells at both the 7- and 14-day time-points. This was observed in both co-culture and conditioned medium experiments (Fig. 2F, J). After 7 days of differentiation with CME11 and CMvm, primary VM cultures comprised Tuj1⫹ or NF⫹ neurons (Balin et al., 1991; Clark and Lee, 1991) and Nestin⫹ neural precursors (Fig. 2D, E, G, J), whereas at 14 days in vitro the majority of cells were NF⫹ (Fig. 2A, B). The differentiation pattern of expanded NSvm and NSE11

612

D. Moses et al. / Neuroscience 154 (2008) 606 – 620

Fig. 2. Rescue of primary DA neurons in co-cultures with NSE11 and NSvm, and respective conditioned medium. Co-cultures of NSvm and NSE11 with primary DA neurons were initiated and fixed on day 7 and day 14 in vitro. Fluorescent double-labeling was carried out to identify cell phenotypes. The number of immunoreactive cells/well and the percentage of immunoreactive cells relative to the total number of Hoescht nuclei were obtained. (A) TH⫹ neurons (red) and NF⫹ neurons (green) in co-cultures of primary DA neurons with NSvm by day 14 in vitro; scale bar⫽100 ␮m. (B) TH⫹ neurons (red) and NF⫹ neurons (green) in co-cultures of primary DA neurons with NSE11 by day 14 in vitro; scale bar⫽100 ␮m. (C) Interface between GFP⫹ NSE11 and GFP-negative primary DA cultures is shown: note single NSE11-derived GFP⫹/TH⫹ cell (arrow) and GFP-negative/TH⫹ cells (arrowheads). (D) TH⫹ neurons (green) and Tuj1⫹ neurons (red) in primary cultures of DA neurons in the presence of CMvm by day 7 in vitro; scale bar⫽40 ␮m. (E) TH⫹ DA neurons (green) and Tuj1⫹ neurons (red) in primary cultures of DA neurons in the presence of NSvm by day 7 in vitro; scale bar⫽40 ␮m. (F) TH⫹ neurons (green) and GFAP⫹ glia (red) in primary cultures of DA neurons in the presence of CMvm by day 7 in vitro. (G): TH⫹ neurons (green) and Nestin⫹ progenitors (red) in primary cultures of DA neurons in the presence of NSvm by day 7 in vitro. Quantification of the total number (H) and the proportion (I) of TH⫹ cells that survived in co-culture with NSvm and NSE11 by day 7 and day 14 in vitro. Data represent mean⫾S.E.M. of four to six independent experiments; n⫽12–14 wells per condition assessed; * P⬍0.05 difference between respective means of NSvm or NSE11 at day 7 and day 14. In control experiments primary DA neurons were cultured in the presence of NS medium alone. TH⫹ cells were absent in control cultures by day 7. (J) Quantitative assessment of neural markers NF, Tuj1, Nestin and GFAP in primary DA cultures following 7 days of differentiation in the presence of CMvm and CME11. Data represent mean⫾S.E.M. of at least three independent experiments; n⫽8 –12 wells per condition assessed; * P⬍0.05 difference between respective means. For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.

D. Moses et al. / Neuroscience 154 (2008) 606 – 620

613

Fig. 3. Rotational asymmetry and anatomical characterization of the nigro-striatal system of 6-OHDA-lesioned mice. Mice received unilateral injection of 6-OHDA into the right SN. Mice were tested for rotational asymmetry 3 weeks later (referred to as post-lesion (PL) on graph) and RS were obtained (for details, see Experimental Procedures). Mice displaying RS⬎100 were selected for further experiments and RS in response to amphetamine challenge tested every 4 weeks. Animals were killed at 16 weeks PL and brains processed for TH and DAT immunolabeling. (A) Individual RS of mice showing that mice with initial RS⬎600 (Group-1) on amphetamine challenge at 3 weeks after 6-OHDA lesion maintained RS on the subsequent challenges. Mice with RS⬍600 (Group-2) had variable RS. (B) Comparison of RS of Group-1 vs. Group-2 demonstrates that Group-1 maintained and even increased rotational asymmetry over the period of 16 wks, whereas RS of Group-2 decreased significantly over the 16 wks (* and #-P⬍0.05 for difference from PL by repeated measures ANOVA followed by Dunnett’s multiple comparison test). (C) Mice in Group-1 had almost complete (⬎96%) loss of TH⫹ cell bodies in the lesioned SN compared with contralateral intact SN. Mice in Group-2 had “partial” lesion of SN displaying loss of about 50% of TH⫹ nigral cell bodies. Group-1 mice also had fewer TH⫹ cells in intact SN compared with Group-2, which is probably due to “spillage” of 6-OHDA to the contralateral SN; * P⬍0.05 and ** P⬍0.01 by ANOVA. (D) Density of DAT⫹ terminals in the dorsal CPu corresponding to lesioned or intact SN of Group-1 and Group-2; ** P⬍0.01 by ANOVA). Note that the density of DAT⫹ terminals in the right dorsal CPu (corresponding to the lesioned SN) of Group-2 was not statistically different from the density of DAT⫹ terminals in the contralateral dorsal CPu, which indicated that reduction of RS in Group-2 mice correlated with the degree of striatal DA innervation.

was not changed by the co-culture conditions (data not shown). Grafts of intact NSE11 or NSvm improve amphetamine-induced rotational behavior in a 6-OHDA model of hemi-Parkinsonism For in vivo study of the NS-driven rescue of injured DA neurons, we engrafted GFP⫹ NS that enable discriminating between graft- and host-derived cells. Xenogeneic transplantation of mouse GFP⫹ NS into a rat 6-OHDA model would require immunosuppression to prevent graft rejection. However, immunosuppressants, such as cyclosporine have been shown to possess intrinsic neuroprotective and in vivo DA trophic properties (Borlongan et al., 1999; Kaminska et al., 2004). Therefore in the present study transplantation of GFP⫹ NS into the 6-OHDA mouse model of PD was conducted, in which case immunosuppression was considered unnecessary as both donors of

GFP⫹ NS and graft recipients were on a C57BL/6 genetic background. Mice received a single unilateral injection of 6-OHDA into the right SNpc. Rodents with unilateral 6-OHDA lesion display rotational asymmetry following systemic administration of amphetamine. Amphetamine-induced rotational asymmetry arises secondary to unilateral DA striatal denervation (Ungerstedt et al., 1974). Rotational asymmetry or rotational score (RS) can be quantified; reduction of RS correlates with restored striatal DA innervation (Friehs et al., 1991; Curran et al., 1993; Hudson et al., 1993). In a preliminary study we established that amphetamine-induced RS ⱖ600/90-min (at 3 weeks following 6-OHDA lesion) identified mice that would maintain or increase RS during long-term monitoring (as determined by quantification of amphetamine-induced RS every 4 weeks, Fig. 3A, B). Postmortem examination was undertaken at 16 weeks after the first amphetamine challenge. It showed that initial RS ⬎600 correlated with anatomical evidence of severe

614

D. Moses et al. / Neuroscience 154 (2008) 606 – 620

Fig. 4. Grafting of intact NS in the midbrain results in behavioral and anatomical improvement in a PD model. Unilateral 6-OHDA-lesioned mice were selected for transplantation based on amphetamine-induced RS at 3 weeks post-6-OHDA lesion (RS at this time pointed is designated PL on the graph). Transplantation of the selected mice was carried out the day after the PL amphetamine challenge. (A) RS of three groups of grafted mice (NSE11-grafted n⫽6; NSvm-grafted n⫽7; control mice n⫽6) in response to amphetamine were obtained at 4, 8, 12 and 22 weeks post-transplantation. Data were compared with two-way ANOVA for repeated measures. This analysis revealed a significant main effect of group [F(2, 80)⫽11.79, P⫽0.0001]. Reduction of RS of NSE11- and NSvm-grafted mice was statistically significant from controls at 12 weeks and 22 weeks (* P⬍0.05, ** P⬍0.01 and *** P⬍0.001; by pairwise comparisons test with Bonferroni correction for multiple comparisons). (B) Number of TH⫹ cell bodies in intact and lesioned SN of NSE11- and NSvm-grafted and control mice (one-way ANOVA; * P⬍0.05 and ** P⬍0.01 relative to controls). (C) Density of DAT⫹ terminals in the respective dorsal CPu innervated by intact or lesioned SN (one-way ANOVA; * P⬍0.01 relative to controls). (D, F, H) Representative midbrain sections of the control, NSvm- and NSE11-grafted mice, respectively that were immunolabeled with TH antibody. Note graft injection tract with respect to TH⫹ cell bodies and the III cranial nerve rootlet; scale bar⫽100 ␮m. Insets in (D) and (F) indicate an interface area between the graft and the host, where ectopic TH⫹ cells with atypical morphology (E and G) were usually located. (I–K) Representative striatal sections immunolabeled with DAT antibody demonstrating DAT⫹ terminals in the dorsal CPu of control mice (I, J) and DAT⫹ terminal in NS-grafted mice (K), scale bar⫽20 ␮m.

nigro-striatal injury as evidenced by profound loss of nigral TH⫹ neurons (16⫾7, TH⫹ cells in the SN) and near complete lack of DAT expression in the dorsal CPu (Fig. 3C, D). Mice with a lower initial RS displayed a progressive decrease of RS on serial amphetamine challenges over 16 weeks (Fig. 3A, B). Postmortem examination of this experimental cohort showed partial lesion of SNpc (592⫾47, TH⫹ cells in the SN) with significant restoration of DAT expression in the dorsal CPu (Fig. 3C, D). This is consistent with our previous observations that partial lesions of SNpc restore striatal DA innervation due to spontaneous compensatory sprouting and this is accompanied by progressive reduction of rotational asymmetry (Stanic et al., 2003; Parish et al., 2005). For all subsequent transplantation studies we therefore used mice that had RS ⱖ600/90min at 3 weeks after 6-OHDA lesion. In the present study a single graft of GFP⫹ NSE11 or NSvm was transplanted in the region of the lesioned SNpc.

Notably, no mortality was observed after NS grafting. Amphetamine-induced RS of both NSE11 and NSvm transplanted mice decreased gradually; by 22 weeks post-grafting, rotational asymmetry was still present but RS reached levels consistent with significant striatal DA re-innervation (Fig. 4A). The magnitude of behavioral improvement was similar in mice grafted with either NSE11 or NSvm (Fig. 4A). Partial rescue of the host nigro-striatal system in NSE11 or NSvm graft recipients In the present study NS were transplanted 3 weeks after intra-nigral administration of 6-OHDA. The putative rescue effect of transplantation on the nigro-striatal system was assessed 22 weeks later by counting TH⫹ neurons in the delineated SN region and by estimating the density of DAT⫹ terminals in the dorsal CPu. Several in vivo studies have shown that intra-nigral administration of 6-OHDA

D. Moses et al. / Neuroscience 154 (2008) 606 – 620

causes early loss of TH expression in nigral DA neurons, which also undergo apoptotic death that may have a slow course of up to 60 days following injection of 6-OHDA (Ljungdahl et al., 1971; Walkinshaw and Waters, 1994; Jeon et al., 1995; Bowenkamp et al., 1996; Bjorklund et al., 1997; Zuch et al., 2000; Stanic et al., 2003). Thus, following intra-nigral 6-OHDA injection it is important to consider the total number of nigral neurons rather than the number of TH⫹ nigral neurons available for rescue at 3 weeks. In this regard, our previous work showed that 18% and 13.6% of total nigral neurons (as determined by Neutral Red staining) survived at 2 and 4 weeks, respectively following intra-nigral injection of 6-OHDA in rats with large lesions (i.e. 75–100%) (Stanic et al., 2003) comparable to the large lesions described in the current study. Thus, the 6-OHDA lesioning paradigm that was used in this study would generate a sufficient number of injured target DA neurons (i.e. surviving/apoptotic nigral neurons that have lost TH-expression following 6-OHDA) and allow a window of opportunity to deliver putative rescue factors (Hoffer et al., 1994; Toledo-Aral et al., 2003). Compared with control group, the number of TH⫹ neurons in the lesioned SN of NSE11- and NSvm-grafted mice was increased significantly by approximately four and seven times, respectively (Fig. 4B, D, F, H). Interestingly, ectopically located TH⫹ cell bodies were consistently detected in the dorsal mesencephalon along the graft injection tract in control mice and NSE11- and NSvm-grafted mice (Fig. 4E, G). Compared with TH⫹ cells in SN, ectopic TH⫹ cells had smaller cell bodies with slender neurite-like structures. Ectopic DAT⫹ cells were absent in the midbrain of NS-grafted mice. The density of DAT⫹ terminals in the right (denervated) dorsal CPu of NSE11-and NSvm-grafted mice was ⬃30% relative to DAT⫹ terminal density in contralateral dorsal CPu, whereas DAT⫹ terminals were almost absent in the denervated dorsal CPu of control mice (Fig. 4C, I, J, K). Ipsilateral intra-striatal injection of the retrograde neuronal tracer fluorogold (FGD) revealed FGD⫹ neurons in the lesioned SN of both NSvm- and NSE11-grafted, but not of control mice (Fig. 5E). Detailed examination revealed ectopic FGD⫹ neurons in the midbrain of both NSE11 and NSvm graft recipients (Fig. 5E). However, the frequency of ectopic FGD⫹ neurons was negligible (only two to four ectopic FGD⫹ neurons/mouse) and these were present in some but not all NSE11 (two mice) and NSvm (one mouse) grafted mice. Are midbrain THⴙ neurons graft- or host-derived? We next determined if midbrain TH⫹ cells were graft- or host-derived based on the expression of enhanced GFP amplified with an anti-GFP antibody. Reliability of this technique for in vivo identification of the grafted cells had been confirmed in a number of previous studies (Ader et al., 2001, 2004; Ourednik et al., 2002). Immunofluorescent double labeling revealed that all TH⫹ cells (SN and ectopic) were GFP-negative in recipients of NSvm grafts and, therefore host-derived (Fig. 5A, C). In recipients of NSE11 grafts all TH⫹ cells in the SN and vast majority of ectopic

615

TH⫹ cells were host-derived; a small proportion of ectopically located TH⫹ cells (⬍2%) were GFP⫹ NSE11 graftderived (Fig. 5B, D). Careful examination of sections did not reveal graft-derived GFP⫹/TH⫹ cells in the region of SN. Long-term survival of NSE11 and NSvm grafts in immuno-competent host and lack of in situ carcinogenesis Long-term survival of NSE11 and NSvm grafts was excellent (Fig. 5A, B, C, H); GFP⫹ cells were densely packed in the graft precluding determination of the graft survival rate (Fig. 5H). The majority of the NS grafts were located above and within the SNpc; GFP⫹ cells migrated only locally in the immediate vicinity of grafts and established an interface with the lesioned SNpc, but did not infiltrate the entire extent of the lesioned SN (Fig. 5A, B). The vast majority of NSE11- and NSvm-derived GFP⫹ cells expressed the glial markers GFAP and NG2 (Fig. 5F, G). These cells may be a source of activity that rescues DA neurons in vitro (Engele et al., 1991; O’Malley et al., 1994; Engele et al., 1996). Immunofluorescent double labeling with GDNF antibody did not reveal GDNF⫹/GFP⫹ cells (data not shown). All grafted mice survived 22 weeks post-transplantation, the pre-determined experimental endpoint. Immunofluorescent double labeling with Ki67 antibody, a marker of proliferating cells, revealed only rare examples of GFP⫹ /Ki67⫹ cells in some graft recipients (data not shown). This would be consistent with lack of uncontrolled cellular proliferation, a hallmark of malignant tumors, confirming previously described benign behavior of fetal-derived NS grafts in vivo. Clusters of round-shaped cells positive for CD45⫹, a marker of hematopoietic lineage cells, were detected along the transplantation tract in control and NSE11- and NSvmgrafted mice (Fig. 5I, J). This was most likely a consequence of local injury accompanied by disruption of the blood– brain barrier and infiltration by CD45⫹ macrophages into the brain parenchyma at the time of transplantation. CD45⫹ cells were juxtaposed to GFP⫹-grafted cells but did not compromise graft survival attesting to the lack of graft rejection. Clusters of CD45⫹ cells did not stain with Ki67 antibody.

DISCUSSION The main finding of this study is that expanded fetalderived NSE11 and NSvm produce factors that rescue mesencephalic DA neurons in vitro and the nigro-striatal system in vivo. In vitro, NSvm- and NSE11-derived factors equipotently rescued primary DA neurons from serum deprivation–induced apoptosis. This result is not surprising given the developmental and anatomical relationship of DA nigral neurons to the striatum and the VM. The absolute number and proportion of TH⫹ neurons in vitro decrease over time. As proliferating DA neuroblasts are not detected, we propose that NS-derived factors promote survival of DA neurons. NS-derived factors also promoted

616

D. Moses et al. / Neuroscience 154 (2008) 606 – 620

Fig. 5. (A–D) Fluorescent double-labeling with TH/GFP antibodies was carried out to identify NSE11 and NSvm graft location and the origin of TH⫹ cells. (A, B) Representative midbrain sections of NSvm and NSE11 portray typical supra-nigral location of grafts; note limited migration of GFP⫹ cells into the host parenchyma, scale bars⫽100 ␮m. (C) An example of NSvm graft– host interface showing TH⫹ cells in the area of SN, scale bar⫽40 ␮m. (D) Ectopic host-derived TH⫹ cells (arrowhead) and ectopic TH⫹/GFP⫹ cells (arrows) derived from NSE11 grafts; scale bar⫽25 ␮m. (E) This representative midbrain sections shows FGD⫹ cell bodies in ipsilateral SN of NSE11-grafted mice 7–10 days following unilateral injection of FGD into the right dorsal CPu. Note ectopic FGD⫹ neurons (arrowheads); scale bar⫽80 ␮m. (F, G) Majority of NSvm and NSE11 differentiate into glial lineage cells expressing GFAP (F) and NG2 (G); scale bars⫽100 and 25 ␮m, respectively. (H) Representative series of midbrain sections from one NSE11-grafted mouse demonstrating GFP⫹ graft survival; scale bar⫽100 ␮m. (I, J) Clusters of CD45⫹ cells within the transplantation cannula tract of control mice; scale bar⫽25 ␮m.

accrual and survival of NF⫹ non-DA neurons that constituted the majority of primary VM cultures by day 14. As NS produce multiple growth factors and cytokines, it remains to be established if rescue of DA neurons was mediated by the same or different factor/s as pan-neuronal survival. Unexpectedly, we found that primary VM cultures were a source of TH-inductive activity, which appeared to be contact-dependent. Induction of TH expression was detected in NSE11derived committed neuronal progenitors, whereas previous work (Ostenfeld et al., 1999) demonstrated that primary VMderived activity was ineffective in inducing TH expression in committed neuronal progenitors derived from striatal NS at

E14. We propose that the greater developmentally determined plasticity inherent in NSE11-derived committed neuronal progenitors may explain this difference (Temple, 2001). We grafted NSvm or NSE11 in the region of the 6-OHDA-lesioned SNpc to test the in vivo efficacy of DA neuronal rescue. Ungrafted 6-OHDA-lesioned mice and control mice displayed loss of TH expression in more than 95% of nigral DA neurons. We found that the number of TH⫹ neurons in the lesioned SN was significantly increased in recipients of NSvm and NSE11 grafts compared with the control group. Grafting of NSvm and NSE11 in-

D. Moses et al. / Neuroscience 154 (2008) 606 – 620

creased the number of TH⫹ cells in the lesioned SN up to 15% relative to the number of TH⫹ cells in the non-lesioned SN. This corresponds to a high rescue efficacy given that only about 14 –18% of nigral DA neurons were anticipated to be available for rescue at 3 weeks after intra-nigral administration of 6-OHDA (Stanic et al., 2003). The TH⫹ cell counts in the SN were comparatively low in this study. As discussed in Experimental Procedures, this is likely to represent an underestimate related to the counting method adopted to accommodate the low cell numbers seen in the context of large nigral lesions. It is of interest that the relatively small numbers of DA neurons could be sufficient to attenuate rotational asymmetry (Meissner et al., 2005). Grafting of NS was also accompanied by a significant preservation of DAT⫹ terminal density in the dorsal CPu (⬃30% of DAT⫹ terminal density in contralateral CPu). This level of striatal innervation was provided by about 15% of the normal number of nigral DA neurons. Previous studies had shown that 30% of nigral DA neurons were sufficient to maintain entirely normal DAT terminal density in the dorsal CPu (Finkelstein et al., 2001). Significantly, the level of nigro-striatal connectivity identified in the current study was associated with significant, albeit partial, correction of rotational asymmetry. Our results indicate that the rescue of DA neurons was mediated by NS-derived factor(s). RT-PCR analysis of expanded NSvm and NSE11 revealed identical expression profile of the selected growth factors and cytokines known to rescue DA neurons from serum deprivation–induced apoptosis in vitro, such as BDNF (Hyman et al., 1991), NT3 (Hyman et al., 1994), PTN (Hida et al., 2003) and FGF-2 (Engele and Bohn, 1991). BDNF, NT3 and FGF-2 also protected DA neurons in vivo against MPTP-induced neurotoxicity or axotomy-induced injury (Chadi et al., 1993; Frim et al., 1994; Hagg, 1998) but in these cases the growth factor was delivered before or at the time of injury; exogenously administered BDNF, NT3 or FGF-2 have not been reported to rescue DA neurons if delivery is delayed. The “bystander” rescue effect of NS-derived activity on DA neurons could be due to a single factor. GDNF effectively rescued lesioned DA neurons in vivo (Bowenkamp et al., 1995; Kirik et al., 2001; Wang et al., 2002) and is currently undergoing clinical evaluation. GDNF is believed to drive the rescue of host DA neurons reported in experimental models of PD deploying grafts of the immortalized murine NSC line c17.2 (Ourednik et al., 2002), normal human NSC (Redmond et al., 2007) and carotid body– derived cells (Toledo-Aral et al., 2003; Villadiego et al., 2005). Our results indicate that the observed NS-derived “bystander” effect is not mediated by GDNF, but is sufficient to rescue serum-deprived DA neurons in vitro and the nigro-striatal system in vivo. This is further corroborated by the lack of nigro-striatal system rescue in the case of control-grafted mice in this study because local injury robustly enhances expression of GDNF and Sonic Hedgehog (SHH) (Appel et al., 1997; Ho and Blum, 1997; Akazawa et al., 2004).

617

Recent discoveries have added new members to the long list of putative DA neurotrophic factors. In this context, FGF-20 has been reported to rescue DA neurons in an in vitro model of 6-OHDA-induced apoptosis (Murase and McKay, 2006). Also, the recently discovered conserved dopamine neurotrophic factor (CDNF), a member of CDNF/MANF (mesencephalic astrocyte-derived neurotrophic factor) family of proteins rescues nigral DA neurons in a 6-OHDA in vivo model of PD (Lindholm et al., 2007). It remains to be shown if the rescue of DA neurons that has been reported in our study is due to molecules such as FGF-20 and/or CDNF/MANF or indeed novel neurotrophic molecules. An alternative hypothesis proposes that application of a single cytokine or trophic factor may not be sufficient to support the long-term survival of DA neurons (Donaldson et al., 2005). Instead, a dynamic cross-talk with the host milieu of the injured substantia nigra may trigger timely release of the diverse trophic molecules by engrafted NSs (Imitola et al., 2004). Nigral DA neurons spontaneously restore nigrostriatal connectivity with normalization of rotational asymmetry over 16 weeks if SNpc lesion size is less than 70% (as determined by TH immunoreactivity), but this mechanism fails with more extensive loss of nigral DA neurons indicating that the local milieu of the extensively lesioned SN lacks intrinsic reparative signals (Stanic et al., 2003). In this context, the failure of survival of the remaining cohort of nigral DA neurons in the PD brain suggests that the same limitations are operative. The present study shows that transplantation of fetal NS is sufficient to rescue injured nigral DA neurons despite the non-permissive milieu of the extensively lesioned murine SN. Given these parallels, our results suggest that a similar approach has the potential to improve PD. Similar to previous reports (Palmer et al., 2001) we also identified ectopic TH⫹ cells with atypical morphology. Our results suggested that ectopic TH⫹ cells did not contribute to attenuation of rotational asymmetry because RS of control mice did not improve despite the presence of ectopic TH⫹ cells. Ectopic TH⫹ cells also lacked significant characteristics of the mesencephalic DA neurons, such as connectivity with striatal target and DAT expression. Despite that, further investigation into the mechanisms that underpin ectopic TH expression in this experimental paradigm is warranted since expression of TH is integral to acquisition of the DA phenotype. We speculate that transplantation injury triggers local invasion of CD45⫹ cells of hematopoietic lineage, such as macrophages, which release inflammatory cytokines (IL-1, for example) that induce TH expression (Ling et al., 1998) in neural progenitors of the adult mesencephalon (Hermann et al., 2006). This may also explain the in vivo induction of TH expression in NSE11-derived committed neuronal progenitors that were always situated within the transplantation tract in the dorsal midbrain, but not in the SN. Previous reports demonstrated migration of grafted NSC to the contralateral SN in the settings of MPTPinduced bilateral SN lesions (Ourednik et al., 2002; Red-

618

D. Moses et al. / Neuroscience 154 (2008) 606 – 620

mond et al., 2007). In our experiments the migration of NS-derived cells was very limited and restricted to the immediate vicinity of the grafts. This could be due to limited migratory capacity of the intact NS progenitors although the paradigm of unilateral 6-OHDA lesion and lack of appropriate signals from contralateral SN may also be relevant (Aboody et al., 2000). Further experiments are warranted to determine if grafting of intact NS prevents migration of the grafted cells (Sanberg, 2007). Finally, we found no evidence of in situ tumorigenesis with long-term (⬃5 months) grafts of intact NSvm and NSE11, confirming the previously reported safety of fetal-derived NPCs.

CONCLUSION In conclusion, our study demonstrated that NS-derived factors rescue DA neurons. We have shown that such factors are present in long-term expanded NS and provide evidence that NS-derived activity is capable of functionally meaningful rescue of nigral DA neurons even in the face of a non-permissive milieu. The findings are relevant because of the potential for novel translational applications (Redmond et al., 2007; Sanberg, 2007). In this context, availability and ease of standardization makes NS cultures especially attractive. Identification of the molecular identity of NS-derived factor/s may ultimately permit design of effective novel treatments for PD. Acknowledgments—RC2 antibody developed by Miyuki Yamamoto was obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by the University of Iowa, Department of Biological Sciences, Iowa City, IA 52242. The project was funded by the NHMRC. J.D. and M.K.H. are NHMRC Practitioner Fellows.

REFERENCES Aboody KS, Brown A, Rainov NG, Bower KA, Liu S, Yang W, Small JE, Herrlinger U, Ourednik V, Black PM, Breakefield XO, Snyder EY (2000) Neural stem cells display extensive tropism for pathology in adult brain: evidence from intracranial gliomas. Proc Natl Acad Sci U S A 97:12846 –12851. Ader M, Schachner M, Bartsch U (2001) Transplantation of neural precursor cells into the dysmyelinated CNS of mutant mice deficient in the myelin-associated glycoprotein and Fyn tyrosine kinase. Eur J Neurosci 14:561–566. Ader M, Schachner M, Bartsch U (2004) Integration and differentiation of neural stem cells after transplantation into the dysmyelinated central nervous system of adult mice. Eur J Neurosci 20:1205– 1210. Akazawa C, Tsuzuki H, Nakamura Y, Sasaki Y, Ohsaki K, Nakamura S, Arakawa Y, Kohsaka S (2004) The upregulated expression of sonic hedgehog in motor neurons after rat facial nerve axotomy. J Neurosci 24:7923–7930. Andres RH, Huber AW, Schlattner U, Perez-Bouza A, Krebs SH, Seiler RW, Wallimann T, Widmer HR (2005) Effects of creatine treatment on the survival of dopaminergic neurons in cultured fetal ventral mesencephalic tissue. Neuroscience 133:701–713. Appel E, Kolman O, Kazimirsky G, Blumberg PM, Brodie C (1997) Regulation of GDNF expression in cultured astrocytes by inflammatory stimuli. Neuroreport 8:3309 –3312. Balin BJ, Clark EA, Trojanowski JQ, Lee VM (1991) Neurofilament reassembly in vitro: biochemical, morphological and immuno-elec-

tron microscopic studies employing monoclonal antibodies to defined epitopes. Brain Res 556:181–195. Barker RA (2006) Continuing trials of GDNF in Parkinson’s disease. Lancet Neurol 5:285–286. Benoit BO, Savarese T, Joly M, Engstrom CM, Pang L, Reilly J, Recht LD, Ross AH, Quesenberry PJ (2001) Neurotrophin channeling of neural progenitor cell differentiation. J Neurobiol 46:265–280. Bjorklund A, Rosenblad C, Winkler C, Kirik D (1997) Studies on neuroprotective and regenerative effects of GDNF in a partial lesion model of Parkinson’s disease. Neurobiol Dis 4:186 –200. Bjorklund A, Stenevi U, Schmidt RH, Dunnett SB, Gage FH (1983) Intracerebral grafting of neuronal cell suspensions. II. Survival and growth of nigral cell suspensions implanted in different brain sites. Acta Physiol Scand 522:9 –18. Borlongan CV, Stahl CE, Fujisaki T, Sanberg PR, Watanabe S (1999) Cyclosporine A-induced hyperactivity in rats: is it mediated by immunosuppression, neurotrophism, or both? Cell Transplant 8:153–159. Bowenkamp KE, David D, Lapchak PL, Henry MA, Granholm AC, Hoffer BJ, Mahalik TJ (1996) 6-Hydroxydopamine induces the loss of the dopaminergic phenotype in substantia nigra neurons of the rat. A possible mechanism for restoration of the nigrostriatal circuit mediated by glial cell line-derived neurotrophic factor. Exp Brain Res 111:1–7. Bowenkamp KE, Hoffman AF, Gerhardt GA, Henry MA, Biddle PT, Hoffer BJ, Granholm AC (1995) Glial cell line-derived neurotrophic factor supports survival of injured midbrain dopaminergic neurons. J Comp Neurol 355:479 – 489. Campos LS (2004) Neurospheres: insights into neural stem cell biology. J Neurosci Res 78:761–769. Chadi G, Moller A, Rosen L, Janson AM, Agnati LA, Goldstein M, Ogren SO, Pettersson RF, Fuxe K (1993) Protective actions of human recombinant basic fibroblast growth factor on MPTP-lesioned nigrostriatal dopamine neurons after intraventricular infusion. Exp Brain Res 97:145–158. Ciliax BJ, Heilman C, Demchyshyn LL, Pristupa ZB, Ince E, Hersch SM, Niznik HB, Levey AI (1995) The dopamine transporter: immunochemical characterization and localization in brain. J Neurosci 15:1714 –1723. Clark EA, Lee VM (1991) Dynamics of mammalian high-molecularweight neurofilament subunit phosphorylation in cultured rat sympathetic neurons. J Neurosci Res 30:116 –123. Curran EJ, Albin RL, Becker JB (1993) Adrenal medulla grafts in the hemiparkinsonian rat: profile of behavioral recovery predicts restoration of the symmetry between the two striata in measures of pre- and postsynaptic dopamine function. J Neurosci 13:3864 – 3877. Deleyrolle L, Marchal-Victorion S, Dromard C, Fritz V, Saunier M, Sabourin JC, Tran Van Ba C, Privat A, Hugnot JP (2006) Exogenous and fibroblast growth factor 2/epidermal growth factor-regulated endogenous cytokines regulate neural precursor cell growth and differentiation. Stem Cells 24:748 –762. Donaldson AE, Marshall CE, Yang M, Suon S, Iacovitti L (2005) Purified mouse dopamine neurons thrive and function after transplantation into brain but require novel glial factors for survival in culture. Mol Cell Neurosci 30:601– 610. Draberova E, Lukas Z, Ivanyi D, Viklicky V, Draber P (1998) Expression of class III beta-tubulin in normal and neoplastic human tissues. Histochem Cell Biol 109:231–239. Drago J, Murphy M, Bailey KA, Bartlett PF (1991) A method for the isolation of purified murine neuroepithelial cells from the developing mouse brain. J Neurosci Methods 37:251–256. Dunnett SB, Björklund A (1992) Neural transplantation: a practical approach. New York: IRL Press, Oxford University Press. Engele J, Bohn MC (1991) The neurotrophic effects of fibroblast growth factors on dopaminergic neurons in vitro are mediated by mesencephalic glia. J Neurosci 11:3070 –3078.

D. Moses et al. / Neuroscience 154 (2008) 606 – 620 Engele J, Rieck H, Choi-Lundberg D, Bohn MC (1996) Evidence for a novel neurotrophic factor for dopaminergic neurons secreted from mesencephalic glial cell lines. J Neurosci Res 43:576 –586. Engele J, Schubert D, Bohn MC (1991) Conditioned media derived from glial cell lines promote survival and differentiation of dopaminergic neurons in vitro: role of mesencephalic glia. J Neurosci Res 30:359 –371. Finkelstein D, Parish CL, Stanic D, Borrelli E, Drago J, Horne MK (2001) The role of dopamine receptors in regulating the size of axonal arbours. In: The international basal ganglia conference (Nicholson L, Faull R, eds), pp VII:313–322. New York: Kluwer Academic/Plenum Publishers. Finkelstein DI, Stanic D, Parish CL, Tomas D, Dickson K, Horne MK (2000) Axonal sprouting following lesions of the rat substantia nigra. Neuroscience 97:99 –112. Franklin KBJ, Paxinos G (1997) The mouse brain in stereotaxic coordinates. San Diego: Academic Press. Freed C, Revay R, Vaughan RA, Kriek E, Grant S, Uhl GR, Kuhar MJ (1995) Dopamine transporter immunoreactivity in rat brain. J Comp Neurol 359:340 –349. Freeman TB, Sanberg PR, Nauert GM, Boss BD, Spector D, Olanow CW, Kordower JH (1995) The influence of donor age on the survival of solid and suspension intraparenchymal human embryonic nigral grafts. Cell Transplant 4:141–154. Friehs GM, Parker RG, He LS, Haines SJ, Turner DA, Ebner TJ (1991) Lesioning of the striatum reverses motor asymmetry in the 6-hydroxydopamine rodent model of parkinsonism. J Neural Transplant Plast 2:141–156. Frim DM, Uhler TA, Galpern WR, Beal MF, Breakefield XO, Isacson O (1994) Implanted fibroblasts genetically engineered to produce brain-derived neurotrophic factor prevent 1-methyl-4-phenylpyridinium toxicity to dopaminergic neurons in the rat. Proc Natl Acad Sci U S A 91:5104 –5108. Goggi J, Theofilopoulos S, Riaz SS, Jauniaux E, Stern GM, Bradford HF (2000) The neuronal survival effects of rasagiline and deprenyl on fetal human and rat ventral mesencephalic neurones in culture. Neuroreport 11:3937–3941. Hagg T (1998) Neurotrophins prevent death and differentially affect tyrosine hydroxylase of adult rat nigrostriatal neurons in vivo. Exp Neurol 149:183–192. Hermann A, Maisel M, Wegner F, Liebau S, Kim DW, Gerlach M, Schwarz J, Kim KS, Storch A (2006) Multipotent neural stem cells from the adult tegmentum with dopaminergic potential develop essential properties of functional neurons. Stem Cells 24:949 –964. Hida H, Jung CG, Wu CZ, Kim HJ, Kodama Y, Masuda T, Nishino H (2003) Pleiotrophin exhibits a trophic effect on survival of dopaminergic neurons in vitro. Eur J Neurosci 17:2127–2134. Ho A, Blum M (1997) Regulation of astroglial-derived dopaminergic neurotrophic factors by interleukin-1 beta in the striatum of young and middle-aged mice. Exp Neurol 148:348 –359. Hoffer BJ, Hoffman A, Bowenkamp K, Huettl P, Hudson J, Martin D, Lin LF, Gerhardt GA (1994) Glial cell line-derived neurotrophic factor reverses toxin-induced injury to midbrain dopaminergic neurons in vivo. Neurosci Lett 182:107–111. Hornykiewicz O (1993) Parkinson’s disease and the adaptive capacity of the nigrostriatal dopamine system: possible neurochemical mechanisms. Adv Neurol 60:140 –147. Hornykiewicz O (2001) Dopamine and Parkinson’s disease. A personal view of the past, the present, and the future. Adv Neurol 86:1–11. Hudson JL, van Horne CG, Stromberg I, Brock S, Clayton J, Masserano J, Hoffer BJ, Gerhardt GA (1993) Correlation of apomorphine- and amphetamine-induced turning with nigrostriatal dopamine content in unilateral 6-hydroxydopamine lesioned rats. Brain Res 626:167–174. Hughes AJ, Daniel SE, Blankson S, Lees AJ (1993) A clinicopathologic study of 100 cases of Parkinson’s disease. Arch Neurol 50:140 –148.

619

Hyman C, Hofer M, Barde YA, Juhasz M, Yancopoulos GD, Squinto SP, Lindsay RM (1991) BDNF is a neurotrophic factor for dopaminergic neurons of the substantia nigra. Nature 350:230 –232. Hyman C, Juhasz M, Jackson C, Wright P, Ip NY, Lindsay RM (1994) Overlapping and distinct actions of the neurotrophins BDNF, NT-3, and NT-4/5 on cultured dopaminergic and GABAergic neurons of the ventral mesencephalon. J Neurosci 14:335–347. Imitola J, Park KI, Teng YD, Nisim S, Lachyankar M, Ourednik J, Mueller FJ, Yiou R, Atala A, Sidman RL, Tuszynski M, Khoury SJ, Snyder EY (2004) Stem cells: cross-talk and developmental programs. Philos Trans R Soc Lond B Biol Sci 359:823– 837. Jensen JB, Bjorklund A, Parmar M (2004) Striatal neuron differentiation from neurosphere-expanded progenitors depends on Gsh2 expression. J Neurosci 24:6958 – 6967. Jeon BS, Jackson-Lewis V, Burke RE (1995) 6-Hydroxydopamine lesion of the rat substantia nigra: time course and morphology of cell death. Neurodegeneration 4:131–137. Jin K, Mao XO, Sun Y, Xie L, Greenberg DA (2002) Stem cell factor stimulates neurogenesis in vitro and in vivo. J Clin Invest 110:311–319. Kaminska B, Gaweda-Walerych K, Zawadzka M (2004) Molecular mechanisms of neuroprotective action of immunosuppressantsfacts and hypotheses. J Cell Mol Med 8:45–58. Kirik D, Georgievska B, Rosenblad C, Bjorklund A (2001) Delayed infusion of GDNF promotes recovery of motor function in the partial lesion model of Parkinson’s disease. Eur J Neurosci 13:1589 – 1599. Lang AE, Lozano AM (1998a) Parkinson’s disease. First of two parts. N Engl J Med 339:1044 –1053. Lang AE, Lozano AM (1998b) Parkinson’s disease. Second of two parts. N Engl J Med 339:1130 –1143. Lin LF, Doherty DH, Lile JD, Bektesh S, Collins F (1993) GDNF: a glial cell line-derived neurotrophic factor for midbrain dopaminergic neurons. Science 260:1130 –1132. Lindholm P, Voutilainen MH, Lauren J, Peranen J, Leppanen VM, Andressoo JO, Lindahl M, Janhunen S, Kalkkinen N, Timmusk T, Tuominen RK, Saarma M (2007) Novel neurotrophic factor CDNF protects and rescues midbrain dopamine neurons in vivo. Nature 448:73–77. Ling ZD, Potter ED, Lipton JW, Carvey PM (1998) Differentiation of mesencephalic progenitor cells into dopaminergic neurons by cytokines. Exp Neurol 149:411– 423. Ljungdahl A, Hokfelt T, Jonsson G, Sachs C (1971) Autoradiographic demonstration of uptake and accumulation of 3H-6-hydroxydopamine in adrenergic nerves. Experientia 27:297–299. Martino G, Pluchino S (2006) The therapeutic potential of neural stem cells. Nat Rev 7:395– 406. Meissner KK, Kirkham DL, Doering LC (2005) Transplants of neurosphere cell suspensions from aged mice are functional in the mouse model of Parkinson’s. Brain Res 1057:105–112. Mizumoto H, Mizumoto K, Shatos MA, Klassen H, Young MJ (2003) Retinal transplantation of neural progenitor cells derived from the brain of GFP transgenic mice. Vision Res 43:1699 –1708. Moses D, Teper Y, Gantois I, Finkelstein DI, Horne MK, Drago J (2006) Murine embryonic EGF-responsive ventral mesencephalic neurospheres display distinct regional specification and promote survival of dopaminergic neurons. Exp Neurol 199:209 –221. Murase S, McKay RD (2006) A specific survival response in dopamine neurons at most risk in Parkinson’s disease. J Neurosci 26:9750 – 9760. Okabe M, Ikawa M, Kominami K, Nakanishi T, Nishimune Y (1997) “Green mice” as a source of ubiquitous green cells. FEBS Lett 407:313–319. O’Malley EK, Sieber BA, Black IB, Dreyfus CF (1992) Mesencephalic type I astrocytes mediate the survival of substantia nigra dopaminergic neurons in culture. Brain Res 582:65–70. O’Malley EK, Sieber BA, Morrison RS, Black IB, Dreyfus CF (1994) Nigral type I astrocytes release a soluble factor that increases

620

D. Moses et al. / Neuroscience 154 (2008) 606 – 620

dopaminergic neuron survival through mechanisms distinct from basic fibroblast growth factor. Brain Res 647:83–90. O’Rourke DA, Sakurai H, Spokes K, Kjelsberg C, Takahashi M, Nigam S, Cantley L (1999) Expression of c-ret promotes morphogenesis and cell survival in mlMCD-3 cells. Am J Physiol 276:F581–588. Ostenfeld T, Horn P, Aardal C, Orpen I, Caldwell MA, Svendsen CN (1999) Mouse epidermal growth factor-responsive neural precursor cells increase the survival and functional capacity of embryonic rat dopamine neurons in vitro. Neuroreport 10:1985–1992. Ostenfeld T, Joly E, Tai YT, Peters A, Caldwell M, Jauniaux E, Svendsen CN (2002) Regional specification of rodent and human neurospheres. Brain Res 134:43–55. Ourednik J, Ourednik V, Lynch WP, Schachner M, Snyder EY (2002) Neural stem cells display an inherent mechanism for rescuing dysfunctional neurons. Nat Biotechnol 20:1103–1110. Palmer MR, Granholm AC, van Horne CG, Giardina KE, Freund RK, Moorhead JW, Gerhardt GA (2001) Intranigral transplantation of solid tissue ventral mesencephalon or striatal grafts induces behavioral recovery in 6-OHDA-lesioned rats. Brain Res 890:86 –99. Parish CL, Finkelstein DI, Drago J, Borrelli E, Horne MK (2001) The role of dopamine receptors in regulating the size of axonal arbors. J Neurosci 21:5147–5157. Parish CL, Nunan J, Finkelstein DI, McNamara FN, Wong JY, Waddington JL, Brown RM, Lawrence AJ, Horne MK, Drago J (2005) Mice lacking the alpha4 nicotinic receptor subunit fail to modulate dopaminergic neuronal arbors and possess impaired dopamine transporter function. Mol Pharmacol 68:1376 –1386. Park KI, Ourednik J, Ourednik V, Taylor RM, Aboody KS, Auguste KI, Lachyankar MB, Redmond DE, Snyder EY (2002) Global gene and cell replacement strategies via stem cells. Gene Ther 9:613– 624. Pluchino S, Quattrini A, Brambilla E, Gritti A, Salani G, Dina G, Galli R, Del Carro U, Amadio S, Bergami A, Furlan R, Comi G, Vescovi AL, Martino G (2003) Injection of adult neurospheres induces recovery in a chronic model of multiple sclerosis. Nature 422:688 – 694. Rafuse VF, Soundararajan P, Leopold C, Robertson HA (2005) Neuroprotective properties of cultured neural progenitor cells are associated with the production of sonic hedgehog. Neuroscience 131:899 –916. Rascol O, Goetz C, Koller W, Poewe W, Sampaio C (2002) Treatment interventions for Parkinson’s disease: an evidence based assessment. Lancet 359:1589 –1598. Redmond DE Jr, Bjugstad KB, Teng YD, Ourednik V, Ourednik J, Wakeman DR, Parsons XH, Gonzalez R, Blanchard BC, Kim SU, Gu Z, Lipton SA, Markakis EA, Roth RH, Elsworth JD, Sladek JR Jr, Sidman RL, Snyder EY (2007) Behavioral improvement in a primate Parkinson’s model is associated with multiple homeostatic effects of human neural stem cells. Proc Natl Acad Sci U S A 104:12175–12180. Reynolds BA, Rietze RL (2005) Neural stem cells and neurospheresre-evaluating the relationship. Nat Methods 2:333–336. Reynolds BA, Weiss S (1992) Generation of neurons and astrocytes from isolated cells of the adult mammalian central nervous system. Science 255:1707–1710. Richardson RM, Broaddus WC, Holloway KL, Fillmore HL (2005) Grafts of adult subependymal zone neuronal progenitor cells rescue hemiparkinsonian behavioral decline. Brain Res 1032:11–22. Rosenblad C, Kirik D, Bjorklund A (2000) Sequential administration of GDNF into the substantia nigra and striatum promotes dopamine neuron survival and axonal sprouting but not striatal reinnervation or functional recovery in the partial 6-OHDA lesion model. Exp Neurol 161:503–516. Rosenblad C, Martinez-Serrano A, Bjorklund A (1998) Intrastriatal glial cell line-derived neurotrophic factor promotes sprouting of spared

nigrostriatal dopaminergic afferents and induces recovery of function in a rat model of Parkinson’s disease. Neuroscience 82:129–137. Sanberg PR (2007) Neural stem cells for Parkinson’s disease: to protect and repair. Proc Natl Acad Sci U S A 104:11869 –11870. Santa-Olalla J, Baizabal JM, Fregoso M, del Carmen Cardenas M, Covarrubias L (2003) The in vivo positional identity gene expression code is not preserved in neural stem cells grown in culture. Eur J Neurosci 18:1073–1084. Sawada H, Ibi M, Kihara T, Urushitani M, Honda K, Nakanishi M, Akaike A, Shimohama S (2000) Mechanisms of antiapoptotic effects of estrogens in nigral dopaminergic neurons. FASEB J 14:1202–1214. Schierle GS, Leist M, Martinou JC, Widner H, Nicotera P, Brundin P (1999) Differential effects of Bcl-2 overexpression on fibre outgrowth and survival of embryonic dopaminergic neurons in intracerebral transplants. Eur J Neurosci 11:3073–3081. Stanic D, Finkelstein DI, Bourke DW, Drago J, Horne MK (2003) Timecourse of striatal re-innervation following lesions of dopaminergic SNpc neurons of the rat. Eur J Neurosci 18:1175–1188. Takeshima T, Johnston JM, Commissiong JW (1994) Mesencephalic type 1 astrocytes rescue dopaminergic neurons from death induced by serum deprivation. J Neurosci 14:4769 – 4779. Temple S (2001) The development of neural stem cells. Nature 414:112–117. Toledo-Aral JJ, Mendez-Ferrer S, Pardal R, Echevarria M, LopezBarneo J (2003) Trophic restoration of the nigrostriatal dopaminergic pathway in long-term carotid body-grafted parkinsonian rats. J Neurosci 23:141–148. Tomac A, Widenfalk J, Lin LF, Kohno T, Ebendal T, Hoffer BJ, Olson L (1995) Retrograde axonal transport of glial cell line-derived neurotrophic factor in the adult nigrostriatal system suggests a trophic role in the adult. Proc Natl Acad Sci U S A 92:8274 – 8278. Uchida N, Buck DW, He D, Reitsma MJ, Masek M, Phan TV, Tsukamoto AS, Gage FH, Weissman IL (2000) Direct isolation of human central nervous system stem cells. Proc Natl Acad Sci U S A 97:14720 –14725. Ungerstedt U, Ljungberg T, Steg G (1974) Behavioral, physiological, and neurochemical changes after 6-hydroxydopamine-induced degeneration of the nigro-striatal dopamine neurons. Adv Neurol 5:421– 426. Villadiego J, Mendez-Ferrer S, Valdes-Sanchez T, Silos-Santiago I, Farinas I, Lopez-Barneo J, Toledo-Aral JJ (2005) Selective glial cell line-derived neurotrophic factor production in adult dopaminergic carotid body cells in situ and after intrastriatal transplantation. J Neurosci 25:4091– 4098. Walkinshaw G, Waters CM (1994) Neurotoxin-induced cell death in neuronal PC12 cells is mediated by induction of apoptosis. Neuroscience 63:975–987. Wang L, Muramatsu S, Lu Y, Ikeguchi K, Fujimoto K, Okada T, Mizukami H, Hanazono Y, Kume A, Urano F, Ichinose H, Nagatsu T, Nakano I, Ozawa K (2002) Delayed delivery of AAV-GDNF prevents nigral neurodegeneration and promotes functional recovery in a rat model of Parkinson’s disease. Gene Ther 9:381–389. West MJ, Gundersen HJ (1990) Unbiased stereological estimation of the number of neurons in the human hippocampus. J Comp Neurol 296:1–22. West MJ, Slomianka L, Gundersen HJ (1991) Unbiased stereological estimation of the total number of neurons in the subdivisions of the rat hippocampus using the optical fractionator. Anat Rec 231: 482– 497. Zuch CL, Nordstroem VK, Briedrick LA, Hoernig GR, Granholm AC, Bickford PC (2000) Time course of degenerative alterations in nigral dopaminergic neurons following a 6-hydroxydopamine lesion. J Comp Neurol 427:440 – 454.

(Accepted 20 March 2008) (Available online 8 April 2008)