|Home | About | Journals | Submit | Contact Us | Français|
Multiple System Atrophy (MSA) is a neurodegenerative disorder characterized by striato-nigral degeneration and olivo-pontocerebellar atrophy. Neuronal degeneration is accompanied by primarily oligodendrocytic accumulation of alpha-synuclein (αsyn) as opposed to the neuronal inclusions more commonly found in other α-synucleinopathies such as Parkinson's disease.
It is unclear how αsyn accumulation in oligodendrocytes may lead to the extensive neurodegeneration observed in MSA, we hypothesize that the altered expression of oligodendrocyte-derived neurotrophic factors by αsyn may be involved. In this context, the expression of a number neurotrophic factors reportedly expressed by oligodendrocytes (Glial-derived neurotrophic factor (GDNF), Brain-derived neurotrophic factor (BDNF) and Insulin-like Growth Factor-1 (IGF1) as well as basic fibroblast growth factor-2 (bFGF2), reportedly astrocytes-derived) were examined in transgenic mouse models expressing human αsyn (hαsyn) under the control of either neuronal (PDGFβ or mThy1) or oligodendrocytic (MBP) promoters.
Whilst protein levels of BDNF and IGF1 were altered in all the αsyn transgenic mice regardless of promoter type, a specific decrease in GDNF protein expression was observed in the MBP-hαsyn transgenic mice. Intracerebroventricular infusion of GDNF improved behavioral deficits and ameliorated neurodegenerative pathology in the MBP-hαsyn transgenic mice.
Consistent with the studies in the MBP-hαsyn transgenic mice, analysis of GDNF expression levels in human MSA samples demonstrated a decrease in the white frontal cortex and to a lesser degree in the cerebellum compared to controls.
These results suggest a mechanism in which αsyn expression in oligodendrocytes impacts on the trophic support provided by these cells for neurons, perhaps contributing to neurodegeneration.
Multiple system Atrophy (MSA) is a sporadic, progressive, neurodegenerative disease which presents with motor abnormalities such as akinesia, ridigidy and postural instability, and is characterized neuropathologically by glial cytoplasmic inclusions (GCI) of alpha-synuclein (αsyn), primarily in oligodendrocytes (Arima et al., 1998; Tu et al., 1998; Wakabayashi et al., 1998b; Wakabayashi et al., 1998a) unlike other alpha-synucleinopathies, such as Parkinson's disease (PD) and Dementia with Lewy bodies (DLB), which are characterized by neuronal aggregates of αsyn. Despite the primarily oligodendrocytic accumulation of αsyn, patients with MSA display considerable neuronal loss in the striatum, cerebellum, brainstem and cortex, accompanied by astrogliosis, microgliosis and myelin loss (Wakabayashi and Takahashi, 2006; Yoshida, 2007). Similar neurodegeneration has been observed in transgenic (tg) mice over expressing human αsyn (hαsyn) under the control of an oligodendrocytic-specific promoter (MBP-myelin basic protein) (Shults et al 2005); these tg mice also develop oligodendrocytic accumulations of αsyn.
Whilst a number of factors other than αsyn accumulation, such as up-regulated apoptotic mechanisms, activated microglia (Probst-Cousin et al., 1998) and mitochondrial dysfunction (Blin et al., 1994; Schulz and Beal, 1994), been linked to cell death in MSA, it is the presence of αsyn that has been most extensively investigated. However, it is as yet unclear how accumulation of oligodendrocytic αsyn can lead to this extensive neurodegeneration. Oligodendrocytes have recently been reported to produce a number of neurotrophic factors (NTFs), including glial-derived neurotrophic factor (GDNF) (Du and Dreyfus, 2002; Wilkins et al., 2003), BDNF (Dai et al., 2001; Dai et al., 2003) and IGF-1 (Wilkins et al., 2001), therefore it is possible that oligodendrocytic αsyn accumulation, as seen in MSA may lead to alterations in the levels of these NTFs.
In order to investigate this hypothesis further we sought to examine the expression of NTFs expressed by oligodendrocytes as well as the reportedly astrocyte-derived basic fibroblast growth factor-2 (bFGF2) (Ferrara et al., 1988), as a control, in tg mice over expressing human αsyn (hαsyn) under the control of either neuronal or oligodendrocytic promoters and in mice deficient for αsyn (αsynKO). We were specifically interested in the NTFs that may be differentially expressed between tg mice with hαsyn under the oligodendrocytic promoter and those with hαsyn under the control of neuronal promoters as we felt this would represent a class of NTF specifically affected by oligodendrocytic expression of αsyn, as seen in MSA, rather than a more general response to αsyn up regulation.
Whilst the levels of many NTFs examined were altered upon α-syn accumulation, only GDNF was specifically reduced in mice expressing α-syn under an oligodendrocytic promoter. These results were confirmed in vitro and GDNF infusion was demonstrated to ameliorate behavioral and neuropathological deficits in the MBP-hαsyn tg mice in comparison to saline-infused MBP-hαsyn tg mice. Consistent with results from the MBP-hαsyn tg mice we demonstrate a reduction in GDNF levels in the white matter of the frontal cortex and cerebellum of human MSA patients by ELISA (enzyme-linked immunosorbent assay) and immunoblot. The results from this study suggest that αsyn accumulation in oligodendrocytes may adversely impact their production of neurotrophic factors such as GDNF, which in turn may impair neuronal health and function, possibly resulting in neurodegeneration.
Mice expressing human αsyn under the control of the MBP promoter (MBP-hαsyn tg) were generated as previously described (Shults et al., 2005); specifically this study used the MBP1 line. The MBP-hαsyn line 1 mice were chosen for this study as they expressed an intermediate level of αsyn expression in comparison to the other lines, they were able to be assessed behaviorally and were breedable. In contrast, other higher expressing lines were less viable and aggressive making them less suitable for this study. The MBP-hαsyn line 1 mice have previously been shown to accumulate αsyn in oligodendrocytes from 3 months of age and to display neuropathological alterations including myelin loss and astrogliosis and behavioral deficits (Shults et al., 2005). Comparisons in the levels of neurotrophic factors were made with mice expressing human αsyn under the control of the neuronal promoters- PDGFβ (platelet-derived growth factor-beta) or mThy1 (Rockenstein et al., 2002) and in mice deficient for αsyn (Stock number: 003692, Jackson Laboratory, Bar Harbor, Maine). Determination of mouse genotype was assessed via genomic DNA that was extracted from tail biopsies and analyzed by PCR amplification, as previously reported (Rockenstein et al., 1995). To investigate the age-related changes in neurotrophic factor expression a total of 15 NTg mice (6 months, n=5, 9 month n=5, 12 months n=5) and 15 MBP-hαsyn (6 months, n=5, 9 month n=5, 12 months n=5) were used.
To investigate the effect of GDNF infusion in mice expressing hαsyn under an oligodendrocytic promoter, a total of 24 mice were used, they received 0.08ug/ul GDNF (Invitrogen) (MBP-hαsyn tg mice, n=6 and non-tg (NTg) mice n=6) or saline (MBP-hαsyn tg mice, n=6 and NTg mice n=6) delivered by intracerebroventricular osmotic minipump (Alzet; Charles River Laboratories) infusion at a rate of 0.25ul/hour for 14 days. The minipump was implanted subcutaneously on the back under light anesthesia as previously described (Rose et al., 2009). All experiments were approved by the animal subjects committee at the University of California, San Diego (UCSD), and were performed according to National Institutes of Health (NIH) recommendations for animal use. Mice underwent behavioral testing in the two weeks following infusion and were sacrificed at the end of this period. Mice used were 8 months old and on a C57/BL6 background.
For the pole test animals were placed head upwards on top of a vertical wooden pole 50cm long and 1cm in diameter. When placed on the pole, animals orient themselves downward and descend the length of the pole. Groups of mice received training that consisted of five trials for each session. For testing animals received five trials and time taken to point downwards (T-turn) and total time to descend (T-Total) were measured.
In order to assess olfactory behavior, mice were placed in a cage in which a food pellet had been hidden under the bedding of the cage at one end. Mice completed 5 trials with the time taken for the mice to find the pellet being recorded each time. Prior to the buried pellet test mice underwent a taste preference test in which they were presented with a number of food pellet options of varying taste and smell – the pellet chosen by the mouse most consistently out of 5 trials was the pellet hidden in the buried pellet test.
Following NIH guidelines for the humane treatment of animals, under anesthesia mice were sacrificed and brains removed. The right hemibrain was immersion-fixed in 4% paraformaldehyde in PBS pH 7.4 and serially sectioned at 40 μm with the Vibratome (Leica, Deerfield, IL) for subsequent analysis of neurodegeneration. The left hemibrain was kept at -80°C for biochemical analysis.
In order to investigate the effects of αsyn on oligodendrocytes in vitro, neuronal precursor cells (NPCs) derived from adult rat hippocampus were differentiated into oligodendrocytes and infected with lenti-virus (LV)-αsyn. Adult rat hippocampal NPCs were cultured routinely for expansion essentially as previously described (Hsieh et al., 2004; Ray and Gage, 2006). Briefly, cells were grown as monolayers on poly-ornithine/laminin coated plates in NPC basal media (DMEM/F-12, 1% N2 supplement, 2mM L-glutamine, and 1% penicillin-streptomycin) supplemented with 20ng/mL fibroblast growth factor-2 (FGF2) (NPC Expansion media). For differentiation into oligodendrocytes, cells were plated onto poly-ornithine/laminin coated plates or coverslips in NPC Expansion media, and transferred the following day (day 0) to NPC Oligodendrocyte Differentiation media (NPC basal media supplemented with 500ng/ml insulin growth factor-1 [IGF-1]). Cells were allowed to differentiate for four days without additional feeding and were infected with LV-αsyn (or empty vector as a control (LV-Control) and left for an additional three days. Following differentiation and infection, cells were lysed in TNE (20mM Tris–HCl, 150mM NaCl, 1mM EDTA, 1% NP-40, 5mM 2-mercaptoethanol, 1× protease inhibitor cocktail (Calbiochem, La Jolla, CA) and 1× phosphatase inhibitor cocktail (Calbiochem, La Jolla, CA) containing sodium orthovanadate) with 1% Triton-X 100 for immunoblot analysis or fixed in 4% paraformaldehyde for immunocytochemistry. Conditioned media was collected from each condition and GDNF levels in this media were assessed via ELISA. KCl (55mM) was used as to enhance GDNF secretion as a positive control, as previously described (Oh-hashi et al., 2009).
To investigate the effects of GDNF in MBP-hαsyn tg mice, 40μm vibratome sections were immunolabeled overnight with antibodies against GDNF (1:250; Santa Cruz), the dendritic marker microtubule-associated protein-2 (MAP2; 1:1000, mouse monoclonal; Chemicon), the neuronal marker NeuN (1:1000, Chemicon) and the astroglial marker glial fibrillary acidic protein (GFAP), the dopaminergic marker tyrosine hydroxylase (TH; 1:200; Chemicon), dopamine transporter (DAT, 1:200; Chemicon) and αsyn (1:1000; Chemicon) followed by incubation with species-appropriate secondary antibodies (1:2000, Vector Laboratories). Sections were transferred to SuperFrost slides (Fisher Scientific, Tustin, CA) and mounted under glass coverslips with anti-fading media (Vector Laboratories). The immunolabeled blind-coded sections were analyzed with the laser scanning confocal microscope (LSCM) (MRC1024, BioRad).
Stereological analysis of the numbers TH and NeuN immunoreactive cells, in the substantia nigra and fronto-parietal cortex respectively, was conducted with the Stereo-Investigator Software (MBF Biosciences) images were collected according to the optical disector method and expressed as cells per unit volume. TH and DAT immunoreactivity levels in the caudate putamen were assessed by obtaining optical density measurements using the Image Quant 1.43 program (NIH) and corrected against background levels, this is presented as ‘corrected optical density’ where the background signal has been subtracted from the optical density measurements.
Protein levels of NTF, GDNF family members, GDNF receptors and total and phosphorylated αsyn (S129) were determined by immunoblot analysis in brain tissue (frozen posterior hemibrain, processed to obtain the cytosolic fraction, as previously described (Shults et al., 2005)) from the MBP-hαsyn (n=5, age=8months), PDGFβ-hαsyn (n=5, age=8months) and mThy1-hαsyn (n=5, age=8months) tg and NTg (n=5, age=8months) mice or cell lysates from oligodendrocytes derived from neuronal precursor cells.
Twenty micrograms of total protein per mouse were loaded onto 4-12% Bis-Tris (Invitrogen) SDS-PAGE gels and transferred onto Immobilon membranes, incubated with two antibodies against GDNF (1:1000; Santa Cruz) and (1:1000;Abcam), BDNF(1:1000 Millipore), IGF1 (1:1000, Abcam), bFGF2 (1:1000, Sigma), Neurturin (1:1000; Millipore), Artemin (1:1000; Abbiotec), RET receptor (1:1000; Abcam), c-RET receptor (1:1000; Chemicon), GFRα3 receptor (1;1000; Chemicon), GFRα1 (1;1000; Abcam), phospho-αsyn (11A5, 1:1000) (Fujiwara et al., 2002) and total αsyn (1:1000; Chemicon). After overnight incubation with primary antibodies, membranes were incubated in appropriate secondary antibodies, reacted with ECL, and developed on a VersaDoc gel-imaging machine (Bio-Rad, Hercules, CA). Anti-beta-actin (1:1000; Sigma) antibody was used to confirm equal loading.
Total RNA was extracted from MBP-hαsyn tg mice and NTg mice, three independent sets of RNA were isolated for each mouse using the RNAeasy kit (QIAGEN, Germantown, MD, USA) as per manufacturer's instructions. All the samples were treated with DNAse I to eliminate genomic DNA contamination. RNA quantification was determined by spectrophotometer readings. The ratio of OD260/OD280 was used to evaluate the purity of the nucleic acid samples and the quality of the extracted total RNA was determined using agarose gel electrophoresis.
For cDNA synthesis, 1 μg total RNA was reverse transcribed using iScript cDNA Synthesis kit (BioRad, Hercules, CA, USA). RT-PCR experiments were performed using the iQ5 Detection System (BioRad, Hercules, CA, USA). Amplification was performed on a cDNA amount equivalent to 25 ng total RNA with 1× iQ SYBRGreen Supermix (BioRad, Hercules, CA, USA) containing dNTPs, MgCl2, Taq DNA polymerase, and forward and reverse primers. PCR reactions were performed on three independent sets of template. Experimental samples and no-template controls were all run in duplicate. The PCR cycling parameters were: 50 °C for 2 min, 95 °C for 10 min, and 40 cycles of 94 °C for 15 s, 60 °C for 1 min. Finally, a dissociation protocol was also performed at the end of each run to verify the presence of a single product with the appropriate melting point temperature for each amplicon. The amount of studied cDNA in each sample was calculated by the comparative threshold cycle method and expressed using mouse actin as an internal control.
An antisence probe was generated against mouse GDNF with the sequence: 5′- 3′: ccaaggaggaactgatctttcgatattgcagcggttcctgtgaatcggcc as previously described (Nosrat et al., 1996) with nucleotide substitutions for the mouse sequence. The probe was labeled with digitoxin using the DIG Oligonucleotide Tailing Kit (Roche, Palo Alto, CA) as per manufacturer's instructions. Paraffin sections were prepared for in situ by de-paraffinization through xylene and graded ethanol from 100% - 50%, incubated in DEPC-PBS (phosphate buffered solution) for 5 mins and then fixed in DEPC-4% paraformaldehyde-PBS. Following acetylation with fresh 0.1M TEA (Triethanolanmine) buffer and acetic anhydride to 0.5% (v/v), slides were washed in 2×SSC. Slides were pre-hybridized at 52°C with DIG Easy Hyb solution (Roche, Palo Alto, CA) for 30 mins and then hybridized for a further 4 hours with DIG Easy Hyb solution (Roche, Palo Alto, CA) plus the DIG-labeled GDNF probe (1:500). Following hybridization slides went though post hybridization washes with 2×SSC and 0.1%×SDS. Slides were then dehydrated and coverslipped. Multiple slides were prepared for double-labeling with cell-type specific antibodies. Slides hybridized with the sense probe were clear.
The concentration of GDNF protein in mouse whole brain homogenate tissue was determined by ELISA, briefly the samples were sonicated in a homogenization buffer (20 mM Tris, pH 8, 137 mM NaCl, 1% Nonidet P-40, 1,7 μg/ml phenylmethylsulfonyl fluoride, 10 μg/ml aprotinin, 1 μg/ml leupeptin and 0.5 mM sodium vanadate) at a tissue concentration of 30 μl/mg. Tissue levels of GDNF was determined in homogenates by ELISA according to the supplier's protocol (Promega, Madison, WI). A similar method was used to determine GDNF levels in the conditioned media from NPC-derived oligodendrocytes infected with LV-αsyn and in the human control and MSA brain samples.
Control and MSA human brain samples from the white matter of the frontal cortex and cerebellum were obtained from the ADRC (Alzheimer's Disease Research Center, University of California, San Diego) (Table 1). Brain samples were homogenized in buffer containing HEPES 1.0mM, Benzamidine 5.0mM, 2-Mercaptoethanol 2.0mM, EDTA 3.0mM, Magnesium Sulfate 0.5mM, Sodium Azide 0.05% (pH to 8.8) and centrifuged at 100K for 1 hour (Ultracentrifuge, Beckkam Coulter) to obtain cytosolic (soluble) and particulate (insoluble, membrane-bound) fractions. The cytosolic fraction was used for immunoblot analysis as described above.
Differences between groups were tested using one and two factor ANOVA with Fisher PLSD posthoc tests. Additional preliminary analysis between control and infused groups was conducted using the unpaired, two-tailed, Student's t-test. All results are expressed as mean +/- SEM.
In order to assess the expression levels NTFs in transgenic mice expressing αsyn under neuronal promoter, oligodendrocytic promoters or αsynKO mice immunoblot analysis was conducted (Fig. 1A). The levels of BDNF protein were reduced in all the lines examined, including the αsynKO mice, compared to NTg mice (Fig. 1A, C), whilst IGF-1 was reduced in the MBP-hαsyn tg mice, the PDGF-β hαsyn tg mice and the αsynKO mice but not the mThy-1-hαsyn tg mice, compared to NTg mice (Fig. 1A, D), bFGF appeared unchanged across the tg lines in comparison to NTg mice (Fig. 1A, E).
The only NTF that was differentially reduced specifically in the MBP-hαsyn tg mice was GDNF. GDNF was identified as a 15kDa band (corresponding to the monomer) and a 30kDa band (corresponding to the homodimer) in the cytosolic fractions of the NTg and various lines of hαsyn tg mice. Levels of GDNF monomer were reduced by 57%, with a similar reduction seen in the dimer in the MBP-hαsyn tg mice when compared to NTg, this was confirmed by the use of two separate GDNF antibodies (Fig. 1A, analyzed in F and G).
Levels of NTFs have been reported to decrease with age (Rudman et al., 1981; Shibasaki et al., 1984; Iranmanesh et al., 1991; Hayashi et al., 2001; Mattson et al., 2004; Tapia-Arancibia et al., 2008; Aleman and Torres-Aleman, 2009), therefore in order to investigate the impact of age on the levels of GDNF and other NTFs, immunoblot analysis of levels of GDNF, BDNF, IGF-1 and bFGF was conducted (Supp. Fig 1). A decrease in BDNF levels was observed with increasing age, from 6, 9 to 12 months of age, in both the NTg and MBP-hαsyn mice, consistent with previous reports (Hayashi et al., 2001), MBP-hαsyn mice displayed reduced BDNF expression levels in comparison to the age-matched NTg mice at each of the three ages examined (Supp. Fig 1). Levels of GDNF remained stable from 6-12 months in the NTg mice, the MBP-hαsyn mice however displayed markedly reduced GDNF expression levels in comparison to the age-matched Ntg mice at each of the three ages examined (Supp. Fig 1).
In order to determine whether levels of other GDNF family members such as Neurturin and Artemin were altered, immunoblot analysis was conduced and showed that levels of these NTFs were similar amongst NTg and MBP-hαsyn tg mice (Fig. 2A, analyzed in B). In contrast, immunoblot analysis of levels of GDNF-family receptors demonstrated an increase in the protein levels of the GFRα1 receptor in the MBP-hαsyn tg mice in comparison to NTg mice (Fig. 2, analyzed in C). Consistent with immunoblot studies, ELISA analysis of GDNF protein levels in whole brain homogenates showed a 60% reduction in GDNF levels in the MBP-hαsyn tg mice (1315pg/ml±104.3) in comparison to NTg mice (3333pg/ml±103.9).
To investigate cell-type specific mRNA expression in the MBP-hαsyn tg mice and NTg mice we conducted ISH and observed GDNF mRNA expression in oligodendrocytes using co-labeling with oligodendrocytic markers MBP (Fig. 3A and B), Galactocerebroside (GalC, Fig. 3C and D) and in astrocytes by co-labeling with Glial fibrillary acidic protein (GFAP, Fig. 3E, F) in both the MBP-hαsyn tg mice and NTg mice. Analysis showed a similar percentage of GDNF positive astrocytes and oligodendrocytes in both the MBP-hαsyn tg mice and NTg mice, with in each case, 3 times more astrocytes expressing GDNF mRNA than oligodendrocytes (Fig. 3I). Consistent with the ISH, quantitative real time PCR (qPCR) showed no difference in the levels of GDNF mRNA between the MBP-hαsyn tg mice and NTg mice.
In order to further examine the relationship between oligodendrocytic αsyn expression and GDNF levels, oligodendrocytes derived from neuronal precursor cells were infected with LV-αsyn and examined for levels of GDNF by immunocytochemistry (Fig. 4A-F) or immunoblot analysis (Fig. 4G).
Differentiation to an oligodendrocytic phenotype was demonstrated by immunoreactivity for the oligodendrocytic marker MBP (Fig. 4A, B) and immunoblot analysis with oligodendrocytic (GalC), astrocytic (GFAP) and neuronal (Tuj1) antibodies (Fig. 4G). Cells infected with LV-αsyn displayed higher levels of αsyn immunoreactivity compared to those infected with LV-control (Fig. 4C, D) and GDNF immunoreactivity was markedly reduced in oligodendrocytes infected with LV-αsyn in comparison to LV-control (Fig. 4E, F). Consistent with the immunohistochemistry, immunoblot analysis demonstrated a reduced expression of GDNF in cells infected with LV-αsyn, as confirmed by two different GDNF antibodies (Fig. 4G). In order to examine the levels of GDNF secreted by the cells under the different conditions, media from each group was taken and analyzed by ELISA (Fig. 4H). Consistent with the immunocytochemistry and immunoblot results, media from cells infected with LV-αsyn expressed less GDNF than that from un-infected cells. Application of KCl (55mM) into the cell culture, an enhancer of GDNF secretion (Oh-hashi et al., 2009) was able to increase GDNF secretion in the un-infected cells by 30%, but had a minimal/no effect on the levels of GDNF secreted into the media by the LV-αsyn infected cells (Fig. 4H). These results imply that in the un-infected cells there maybe some GDNF ‘reserve’ that is able to be released upon KCl administration, however the LV-αsyn infected cells lack this reserve and are unable to increase secretion upon KCl administration, perhaps due to the overall decreased expression of GDNF in these cells.
Taken together the results from the MBP-hαsyn tg mice and the NPC-derived oligodendrocytes suggest that over expression of αsyn effects the expression levels of GDNF which may in turn also alter its secretion characteristics.
Although a number of NTFs were disturbed in the MBP-hαsyn tg mice we chose to focus on GDNF as it was the only NTF that appeared to be specifically altered in the MBP-hαsyn tg mice. Having established that the levels of GDNF were reduced in the MBP-hαsyn tg, we sought to investigate whether GDNF infusion might revert the behavioral and neurodegenerative alterations observed in the MBP-hαsyn tg mice. For this purpose, mice received 0.08ug/ul GDNF delivered by osmotic pump infusion over a period of 14 days and underwent behavioral testing in the two weeks following infusion.
Pole test analysis, designed to investigate motor behavior (Matsuura et al., 1997), showed a significant increase in T-Total and T-Turn times in the saline infused MBP-hαsyn tg mice in comparison to saline infused NTg controls (Fig. 5A, B) as has been previously reported (Shults et al., 2005; Ubhi et al., 2009). GDNF infusion significantly decreased T-Total and T-Turn times in the MBP-hαsyn tg mice in comparison to the saline infused MBP-hαsyn tg mice (Fig. 5A, B). GDNF infusion had no effect on the T-Total and T-Turn times of NTg mice.
In the buried pellet test, designed to assess olfactory behavior, saline infused MBP-hαsyn tg mice took significantly longer to find the buried food pellet than the saline infused NTg control mice, indicating an olfactory deficit in these animals (Fig. 5C). GDNF infusion significantly lowed the time taken by the MBP-hαsyn tg mice to find the pellet in comparison to saline infused MBP-hαsyn mice, bringing their time closer to saline infused NTg control mice (Fig. 5C). GDNF infusion had no effect on the time taken by the NTg mice to find the pellet.
Immunohistochemical analysis was performed to assess the effects of GDNF infusion on neuropathology in the MBP-hαsyn tg mice. GDNF immunoreactivity was detected in the neocortex and white matter tracts of saline infused NTg and MBP-hαsyn tg mice, the antibody also recognized cells with oligodendroglial and astroglial characteristics along the white matter tracts in the corpus callosum and striatum (Fig. 6A-D, inserts at higher magnification). Consistent with immunoblot analysis, immunohistochemical studies showed a significant reduction in the levels of GDNF immunoreactivity in the saline infused MBP-hαsyn tg mice in comparison to the saline infused NTg control mice (Fig. 6A, C, analyzed in E). GDNF infusion significantly increased levels of GDNF immunoreactivity in both MBP-hαsyn tg and NTg mice (Fig. 6B, D, analyzed in E). Immunohistochemical analysis of GFAP showed no differences between saline- or GDNF infused MBP-hαsyn tg mice and NTg mice (Fig. 4F-I, analyzed in J).
Immunohistochemical analysis of dendritic pathology, as evidenced by MAP-2 immunoreactivity, demonstrated a significant reduction of MAP-2 immunoreactive neuropil in the saline infused MBP-hαsyn tg mice in comparison to the saline infused NTg mice (Fig. 7A, C), GDNF infusion ameliorated the dendritic pathology as evidenced by MAP-2 immunoreactive neuropil in the MBP-hαsyn tg mice, bringing it into line with NTg levels (Fig. 7A, C, D, analyzed in E). GDNF infusion had no significant effect on the levels of MAP-2 immunoreactive neuropil in the NTg mice. Consistent with previous studies (Shults et al 2005), the saline infused MPBP-hαsyn Tg mice displayed a significant reduction in the number of NeuN-immunoreactive neurons in the neocortex compared to saline infused NTg (Fig. 7F, H), which was improved upon GDNF infusion (Fig. 7F, G, I and analyzed in J). Taken together these studies suggest that the reduced levels of GDNF observed in MBP-hαsyn tg mice, in concert with alterations in other NTFs, plays a role in the neuropathological alterations present in these tg mice and that GDNF infusion is able to alleviate the neuropathological and behavioral deficits observed in these tg mice.
Analysis of TH immunoreactivity in the caudo-putamen (Fig. 8A-D, analyzed in E) and substantia nigra (Fig. 8A-I, analyzed in J) demonstrated a significant decrease in TH-immunoreactive fibers and cells in these regions in the saline infused MBP-hαsyn mice in comparison to the saline infused NTg mice. GDNF infusion increased TH immunoreactivity in fibers in the caudo-putamen (Fig. 8A, D) and cells in the substantia nigra (Fig. 8A, I) in the MBP-hαsyn Tg mice in comparison to saline infused MBP-hαsyn Tg mice. These effects of GDNF on TH immunoreactivity in regions involved with motor behavior are consistent with previous reports (Sun et al., 2004; Pascual et al., 2008) and with the pole test results presented here. Analysis of DAT immunoreactivity in the caudo-putamen (Fig. 8K-N, analyzed in O) demonstrated a significant decrease in DAT-immunoreactive fibers in the saline infused MBP-hαsyn mice in comparison to the saline infused NTg mice (Fig. 8K, M analyzed in O) and GDNF infusion increased DAT immunoreactivity in fibers in the caudo-putamen in the MBP-hαsyn Tg mice in comparison to saline infused MBP-hαsyn Tg mice (Fig. 8M, N analyzed in O).
In order to investigate the effects of GDNF infusion on protein levels of αsyn, immunohistochemistry and immunoblot were conducted. Consistent with previous reports (Shults et al., 2005; Ubhi et al., 2009), immunohistochemical analysis demonstrated abundant αsyn immunoreactive oligodendrocytes in the saline infused MBP-hαsyn tg mice in comparison to saline infused NTg mice (Fig. 9A, E, C, G, analyzed in I). GDNF infusion had no effect on α-syn levels in the MBP-hαsyn tg or NTg mice, (Fig. 9B, F, D, H, analyzed in I).
Immunoblot analysis of the MBP-hαsyn tg mice showed a significant increase in phosphorylated (Supp. Fig. 2A and C) and total (Supp. Fig. 2A and D) αsyn levels in the saline infused MBP-hαsyn tg mice in comparison to saline infused NTg mice and, consistent with the immunohistochemistry, immunoblot analysis showed no effect of GDNF infusion on levels of phosphorylated or total αsyn in either the MBP-hαsyn tg or NTg mice. These results indicate that the beneficial effects of GDNF infusion on motor and olfactory behavior and on TH-cell survival occur in an αsyn-independent manner.
Having established a reduction in GDNF levels in the MBP-hαsyn mice, we were interested in investigating whether a similar deficit was apparent in human MSA samples. In order to examine GDNF levels in humans, immunoblot and ELISA analysis was conducted (Fig. 10) on samples from the white matter of the frontal cortex and cerebellum in human patients, key areas implicated in MSA pathology (Armstrong et al., 2007; Brenneis et al., 2007; Chang et al., 2009). Immunoblot analysis demonstrated a significant decrease of 50% in the levels of GDNF in the white matter of the frontal cortex of the MSA patients in comparison to the controls (Fig. 10A, quantified in B), in the cerebellum however, whilst there was a greater variability in the expression of GDNF, the overall trend also seems to suggest reduced levels in the MSA patients in comparison to controls (Fig. 10A, quantified in B). GDNF levels measured by ELISA were consisted with the immunoblot results (Fig. 10C).
The present study sought to investigate the hypothesis that oligodendrocytic accumulation of αsyn, as observed in MSA, may disrupt expression of oligodendrocyte-derived NTFs. Our results demonstrate that whilst levels of BDNF and IGF1 were disrupted in tg mice over expressing αsyn under both neuronal and oligodendrocytic promoters, only GDNF was specifically reduced in the MBP-hαsyn mice that accumulate αsyn in oligodendrocytes. This finding was corroborated in vitro with oligodendrocytes derived from NPCs and infected with LV-αsyn. The use of KCl as an inducer of GNDF secretion in the NPC-derived oligodendrocytes provided important information as to the impact of the GDNF deficit in LV-αsyn infected cells. Whilst the un-infected cells were able to increase GDNF secretion upon KCl administration, indicating the presence of a GDNF ‘reserve’, the LV-αsyn-infected cells were unable to do so, this suggests that LV-αsyn expression not only reduces the expression level of GDNF but also interferes with the secretion of GDNF by these cells.
Consistent with the results from the MBP-hαsyn tg mice and the NPC-derived oligodendrocytes, examination of GDNF levels in human MSA patients also demonstrated a reduction in GDNF levels, indicating that GDNF may also play an important role in MSA in humans.
Collectively these results point towards the possibility that oligodendrocytic accumulation of αsyn in MSA may interfere with the expression of NTFs and disrupt the neurotrophic support provided by the oligodendrocytes to the neurons, resulting in altered neuronal function and survival.
Several NTFs such as NGF, BDNF and GDNF itself have been widely reported to be altered in a number of neurodegenerative disorders (Siegel and Chauhan, 2000; Capsoni and Cattaneo, 2006; Zuccato and Cattaneo, 2009) and many studies have looked at NTF replacement strategies in an effort to stave of neuronal dysfunction and death in these disorders with a number reporting beneficial effects (Lad et al., 2003; Peterson and Nutt, 2008; Ramaswamy et al., 2009; Sadan et al., 2009; Saragovi et al., 2009). Whilst we report alterations in a number of NTFs in the MBP-hαsyn tg mice, we chose to focus on GDNF as its role in MSA remains elusive, unlike BDNF (Kawamoto et al., 1999; Nishimura et al., 2005) and bFGF-2 (Sarchielli et al., 2008) which have been reported to be altered in MSA. We were also interested in GDNF in relation to its role in olfactory function, an early deficit in MSA and other α-synucleinopathies (Abele et al., 2003) and its reported ability to restore motor function (Gill et al., 2003; Slevin et al., 2005). Though we chose to focus on GDNF we appreciate that augmentation of levels of the other NTFs may also have had an equally beneficial effect.
In addition to its expression by oligodendrocytes (Du and Dreyfus, 2002; Wilkins et al., 2003), GDNF has also been reported in cortical neurons (Schaar et al., 1993; Choi-Lundberg and Bohn, 1995; Pochon et al., 1997) and astrocytes (Ho et al., 1995), it therefore could be argued that the loss of GDNF seen in the MBP-hαsyn tg mice may represent a loss from one of these other sources however given the lack of cortical astrogliosis, which would indicate the loss of surrounding cortical neurons (and hence a possible loss of neuronally-derived GDNF) and the lack of a reduction in GFAP immunoreactivity in the MBP-hαsyn Tg mice, which would indicate reduction in astrocyte number (and hence astrocyte-derived GDNF), the most parsimonious explanation is that the loss of GDNF observed in these mice is indeed that of oligodendrocytic origin. We also demonstrate that the expression of GDNF is reduced at a protein level, with GDNF mRNA as assessed by ISH and qPCR showing no differences between MBP-hαsyn tg mice and NTg mice. This would indicate that the effects of αsyn on GDNF are post-transcriptional, perhaps involving the stability or degradation of GDNF; these effects are potentially very interesting and merit further characterization. This loss is specific to GDNF, with other family members being unaltered in the MBP-hαsyn tg mice whilst analysis of GFR demonstrated a specific alteration only in levels of GFRα1, the GDNF-specific receptor.
In an effort to assess the pathological importance of the GDNF deficit an infusion of GDNF was performed and the mice were assessed behaviorally and biochemically. The MBP-hαsyn tg mice have been extensively characterized and shown to display motor behavioral deficits as assessed by the pole test (Shults et al., 2005; Ubhi et al., 2009), a test of basal ganglia function (Matsuura et al., 1997). The results here demonstrate an attenuation of the motor deficits in the MBP-hαsyn tg mice upon GDNF infusion, these results were consistent with the increase in caudo-putamen and substantia nigra TH immunoreactivity observed in the GDNF infused MBP-hαsyn tg mice compared to saline infused MBP-hαsyn tg mice and with previous reports of the effects of GDNF on TH-positive cells (Boger et al., 2006; Emborg et al., 2008; He and Ron, 2008; Pascual et al., 2008) and a previous study showing behavioral improvement in a mouse model of MSA that had been treated with Rasagiline (Stefanova et al., 2008). Whilst Rasagiline, a drug used in the treatment of PD, has been reported to act via GDNF and BDNF induction (Stefanova et al., 2008), this study is the first to report a behavioral attenuation in a tg model of MSA that can be directly attributed to an increased expression of GDNF.
Additional behavioral deficits, specifically of olfactory function, were assessed via the buried pellet test. Here again GDNF infusion attenuated the behavioral deficit by significantly reducing the time taken by the MBP-hαsyn tg mice to locate the buried food pellet in comparison to saline infused MBP-hαsyn tg mice. These results are important for a number of reasons; firstly this study is the first to demonstrate the olfactory deficit in the MBP-hαsyn tg mice and secondly olfactory dysfunction is widely reported by sufferers of a number of alpha-synucleinopathies including MSA (Nee et al., 1993; Abele et al., 2003; Hawkes, 2006), additionally, MSA patients have GCI in their olfactory bulbs and neuronal loss in the anterior olfactory nucleus (Kovacs et al., 2003). These results are consistent with studies reporting the ability of GDNF to promote the migration of olfactory ensheathing cells, glial cells that ensheath the olfactory nerves (Cao et al., 2006). Further investigation of the expression of GDNF in the olfactory bulbs of the MBP-αsyn and the role it may play in the observed olfactory deficit is necessary to fully understand the mechanisms that may underlie this symptom.
GDNF has been extensively investigated in relation to PD (Barker, 2006; Chen et al., 2008; Emborg et al., 2008; Hong et al., 2008). Initially purified in 1993 as a growth factor promoting embryonic midbrain dopaminergic neurons survival (Lin et al., 1993; Lin et al., 1994) it has since been shown to being necessary for the survival of adult catecholaminergic neurons (Pascual et al., 2008) as well as a trophic factor for spinal motorneurones (Henderson et al., 1994) and central noradrenergic neurons (Arenas et al., 1995). Though the mechanisms through which αsyn expression alters GDNF levels remain unclear, the results from this study demonstrate no alteration of GDNF mRNA levels in the MBP-hαsyn tg mice in comparison to the NTg mice, suggesting that whatever effect αsyn is having, it is post-transcriptionally located. The apparent up regulation of the GFRα1 receptor subunit in the MBP-hαsyn mice is also interesting as it suggests a potential compensatory mechanism in these mice. Further characterization of the GDNF signaling pathway, secretion mechanisms and inter-connected pathways is a key area of future investigation.
Despite the well known role of GDNF in PD (Barker, 2006; Patel and Gill, 2007; Yasuhara et al., 2007; Hong et al., 2008) the role of this NTF in MSA has remained unclear, collectively the results from this study suggest a pathway of communication between oligodendrocytes and neurons, involving the production of NTFs, which may become perturbed upon oligodendrocytic αsyn accumulation as observed in MSA and suggest that GDNF, in combination with other NTFs such as BDNF, may have a wider therapeutic relevance than previously supposed.
This work was supported by NIH grants AG 18440, NS 044233, AG 10435 and AG 022074.