To simulate PD-like neurodegeneration, primary vMB cultures of three to five DIV were exposed to the neurotoxic DA analogue 6-OHDA. RXR ligands were added three hours prior to the toxin, and after 20 minutes of 6-OHDA exposure the medium and ligands were replaced to avoid unspecific stress from oxidized 6-OHDA. After additional 24 hours, cells were fixed, stained and scored. 6-OHDA produced clear signs of neurodegeneration in the TH-positive population, with evident neuronal loss, leaving cell debris and damaged neurites (Figure 1A). The degeneration was dose dependent (ANOVA, p < 0.01, F 12.6) (Figure 1B). 15 μM 6-OHDA reduced the number of DA cells by 60% compared to control (p < 0.01). Simultaneous exposure of the RXR ligand LG268, rescued the DA neurons from degeneration to 95% of control (p < 0.001, 6-OHDA vs. LG268 + 6-OHDA) (Figure 1C). RXR is ubiquitously expressed and it heterodimerizes with several NR subtypes. However, RXR-ligand dependent neurotrophic effects have been shown to be mediated through Nurr1 6. Here, the positive ligand effects after 6-OHDA exposure were only seen for the Nurr1 expressing DA neurons, whereas no neuronal rescue could be detected for the general neuronal population in the culture, as determined by scoring TuJ1-positive cells (Figure 1D). To further relate the trophic effects to Nurr1, we used the RXR ligand XCT, which is selective for the Nurr1-RXR heterodimer in the dose applied (1 μM, see 6). Indeed, XCT rescued vMB DA neurons to 103% of control (p < 0.001, 6-OHDA vs. XCT + 6-OHDA) (Figure 1C). Without exposure to toxic stress, none of the ligands alone had any effect on cell number at three to five DIV (data not shown).

To study whether the neurotrophic activities of the RXR ligands were general or rather selective for specific cellular challenges, we studied their effects on several types of stressors and toxins. Primary vMB DA neurons of three to five DIV were stressed by hypoxic culture conditions (0-1% O₂, 5% CO₂). Ligands were added three hours prior to the hypoxic insult and kept throughout the experiment. Cells were fixed, stained and scored after different time intervals. In control culture conditions there is no decrease in the number of DA cells over time. However, hypoxia induced a time dependent decrease in DA cells (ANOVA, p < 0.001, F 55.2) (Figure 2A). Simultaneous exposure of LG268 during 24 and 36 hours in hypoxic conditions partly rescued the DA cells, giving a 30% (p = ns) and 40% (p < 0.05) increase, respectively, compared to the untreated cells (Figure 2B). Further, the excitotoxic glutamate analogue KA and the oxidative agent H₂O₂ were applied to primary vMB neurons of three to five DIV. KA as well as H₂O₂ exposure resulted in dose dependent degeneration of the DA neurons (KA, ANOVA, p < 0.001, F 277.9) (H₂O₂, ANOVA, p < 0.001, F 79.4) (Figures 2C and 2D). However, addition of the RXR ligands LG268 or XCT provided no neuronal rescue at any of the doses tested. This implies that the ligands are acting on specific cell death mechanisms, induced by 6-OHDA and hypoxia, but not on the calcium dependent excitotoxicity caused by KA or the pure oxidative stressor H₂O₂.

We further used mESC-derived DA cells as platforms for our evaluation of RXR ligands. These cells represent great progress for in vitro studies, since they provide a very high number of the correct cell type and homogeneous cultures. Stable NesE-Lmx1a mESCs (Lmx1a cells) were induced for neuronal differentiation as monolayer cultures in medium supplemented with the DA instructive cues Sonic Hedgehog (Shh) and Fibroblast growth factor (FGF) 8 28. After 18 days of differentiation, cells have a mature phenotype with long extended neuritis and expression of relevant DA cell proteins including TH (Figure 3, see also 28). Cells were given RXR ligands for three hours and were then exposed to 6-OHDA for another three hours. Subsequently, media was changed and fresh ligands were added. Cultures were kept for additional 48 hours before fixation, staining and scoring. 6-OHDA produced a dose dependent toxicity, with reduced number of TH-positive cells and degenerate morphology (ANOVA, p < 0.001, F 83.9) (Figures 3A and 3C). As for the primary vMB neurons, simultaneous exposure of 6-OHDA stressed cultures to LG268 completely rescued the DA neurons from 55% (p < 0.01) reduction to 100% of control (p < 0.01, 6-OHDA vs. LG268 + 6-OHDA) (Figures 3B and 3C). No ligand-effect could be detected in the TuJ1⁺/TH^- neuronal population (data not shown). Also the selective Nurr1-RXR ligand XCT fully rescued the vMB DA neurons from 6-OHDA-induced degeneration (p < 0.001, 6-OHDA vs. XCT + 6-OHDA) (Figure 3B), suggesting an important role for Nurr1. Treatment of the cultures with RXR ligands without prior stressful insult did not affect the cell number (data not shown). Thus, the mESC-derived DA neurons reacted similar as primary neurons to the 6-OHDA stress and the RXR ligands. However, the much higher number of DA cells in culture substantially increased the power of the study. Furthermore, once the proper mESC line was established and the differentiation protocols were refined, the mESCs offered more accessible cultures than those of primary neurons, allowing faster and more experiments.

PD inflicts considerable suffering for the affected and their families, and it constitutes a great cost for the society. Today, the opportunities for pharmacological treatment are meager and mainly consist of substitution with the DA precursor Levodopa. However its medical potential decreases severely over time 30. The development of neuroprotective or neuroregenerative therapies would provide great benefits. As such, glial cell line-derived neurotrophic factor (GDNF) has been tested in several clinical trials against PD 3132333435. Unfortunately, lack of efficacy and potential hazardous side effects has lessened the initial enthusiasm, and future development of GDNF-based drug therapies seems uncertain. Thus, there is a high need for alternative strategies to treat PD.

NR ligands are small and lipophilic, and they easily pass the blood-brain-barrier, providing excellent targets for the pharmaceutical industry. Indeed, the characterization of novel NR-mediated signaling pathways during the last decade has resulted in development of novel drugs used in the treatment of metabolic disease and cancer. Specifically, RXR ligands have limited toxicity in humans 36 and selective RXR ligands would minimize the risk of adverse effects. Our findings that RXR activity leads to neuroprotection 6 provide novel possible approaches to PD research. The neuroprotective effect is mediated, at least partly, by the Nurr1-RXR dimer. Here, we further show that RXR ligands provide neurotrophic support after stress induced by the PD modeling toxin 6-OHDA. The effects are selective for the Nurr1-expressing DA cells and do not affect total neuronal number in the cultures. Interestingly, neuroprotective effects are also seen after hypoxia-induced neurodegeneration. No protective effects are seen after excitotoxic or pure oxidative stress induced by KA and H₂O₂, respectively. Thus, the effects of ligand activation is not generally neuroprotective, but rather likely to affect individual cell death pathways.

All experiments were performed in accordance with guidelines from the Swedish National Board for Laboratory Animals. Primary vMB cultures were obtained as previously described 6. Briefly, vMB from rat embryos at stage E14.5 were dissected, mechanically dissociated and plated on poly-D-lysine coated 12 or 24 well plates in serum free medium (N2) consisting of a 1:1 mixture of MEM (Gibco, UK) with 15 mM Hepes buffer (Gibco, UK) and Ham's F12 medium (Gibco, UK). The mixture was supplemented with 6 mg/ml glucose, 1 mg/ml bovine serum albumine, 5 μg/ml insulin, 100 μg/ml transferrin, 60 μM putrescine, 20 nM progesterone, 30 nM selenium and 1 mM glutamine (Sigma). Cultures were incubated for three to five DIV before treatment. Cultures were kept in 37°C, 5% CO₂ and 99% humidity unless otherwise stated.

DA cells were derived from E14.1 mESCs as previously described 2728. Briefly, mESCs were propagated in feeder free conditions in DMEM (Invitrogen) supplemented with 2000 U/ml LIF (Chemicon), 9% KSR, 3% FBS, 0.1 mM non-essential amino acids, 1 mM pyruvate (Invitrogen) and 0.1 μM β₂-mercaptoethanol (Sigma). For generation of stable ESC lines, 2 × 10^6 cells were nucleofected with 7 μg linearized NesE-mLmx1a-PGK-neo vector according to protocol (mouse ESC nucleofector kit, Amaxa biosystems Gmbh, Koeln, Germany), selected with G418, replated on gelatinized dishes and induced to differentiate in N2B27 differentiation medium 37 supplemented with 20 ng/ml bFGF, 100 ng/ml FGF8 and 70 nM Shh. Cells were incubated for 18 days before treatment. Cultures were kept in 37°C, 5% CO₂ and 99% humidity throughout all experiments.

Experiments were performed in triplicates. The ligands (stock solutions in DMSO; LG268 and XCT, kindly provided by Dr. Mark Leibowitz at Ligand Pharmaceuticals and Dr. Peter Ordentlich at X-ceptor pharmaceuticals, respectively), were diluted to working dilutions in culture medium and added to the cultures for two to three hours prior to neurodegenerative stressors. Final concentrations were 0.1 and 1 μM for LG268 and XCT, respectively. The DA analogue 6-OHDA was given to the vMB primary neuron cultures at 5-15 μM for 20 minutes before medium and ligand replacement. Hypoxic stress was induced by placing cultures in a modular incubator chamber (Billups-Rothenberg Inc., Del Mar, CA) at 37°C filled with 5% CO₂ and 0-1% O₂ (balanced with N₂) from one to 36 hours. The excitotoxic glutamate analogue KA and the oxidative agent H₂O₂ were added over night, in the range of 100 to 500 μM and 80 to 110 μM, respectively. To stress mESC-derived DA cells, 150 to 500 μM 6-OHDA in 0.1% ascorbic acid were added three hours before medium and ligand exchange. The cultures were left for another 36 hours in the incubator.

Paraformaldehyde fixed cultures were incubated overnight with TH (1:1000, Pel-Freez, Arkansas), TuJ1 (1:1000, Babco) antiserum in PBS containing 5% fetal calf serum and 0.3% triton X-100. Following rinses, cultures were incubated with FiTC- and Cy3 conjugated secondary antibodies (Jackson, ImmunoResearch, West Grove, PA) for direct detection or with biotinylated secondary antibodies followed by detection of immuno-staining using the ABC immunoperoxidase kit from Vector (Buringame, CA).

Analysis, imaging and cell counting were performed on Eclipse E1000M and Eclipse TE300 microscopes (both Nikon) coupled to the Spot2 camera (Diagnostic Instruments, Sterling Heights, MI). Scoring was performed by cell counting, and counts were made blind to avoid observation bias. Statistical analyses were performed by one-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test when appropriate.