Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Neurosci. Author manuscript; available in PMC 2013 February 21.
Published in final edited form as:
PMCID: PMC3578394

Interactions of Wnt/β-Catenin Signaling and Sonic Hedgehog Regulate the Neurogenesis of Ventral Midbrain Dopamine Neurons


Signaling mechanisms involving Wnt/β-catenin and sonic hedgehog (Shh) are known to regulate the development of ventral midbrain (vMB) dopamine neurons. However, the interactions between these two mechanisms and how such interactions can be targeted to promote a maximal production of dopamine neurons are not fully understood. Here we show that conditional mouse mutants with region-specific activation of β-catenin signaling in vMB using the Shh-Cre show a marked expansion of Sox2-, Ngn2-, and Otx2-positive progenitors, but perturbs their cell cycle exit and reduces the generation of dopamine neurons. Furthermore, activation of β-catenin in vMB also results in a progressive loss of Shh expression and Shh target genes. Such antagonistic effects between the activation of Wnt/β-catenin and Shh can be recapitulated in vMB progenitors and in mouse embryonic stem cell cultures. Notwithstanding these antagonistic interactions, cell type-specific activation of β-catenin in the midline progenitors using the Th-IRES-Cre leads to increased dopaminergic neurogenesis. Together, these results indicate the presence of a delicate balance between Wnt/β-catenin and Shh signaling mechanisms in the progression from progenitors to dopamine neurons. Persistent activation of β-catenin in early progenitors perturbs their cell cycle progression and antagonizes Shh expression, whereas activation of β-catenin in midline progenitors promotes the generation of dopamine neurons.

Keywords: Wnt, β-Catenin, Sonic Hedgehog, Dopaminergic Neurons, Neurogenesis, Stem Cells, Neuronal Progenitor Cells


The developing ventral midbrain (vMB) in vertebrates contains a neurogenic niche that is enriched with progenitor cells for dopamine (DA) neurons (Bjorklund and Lindvall, 1984). Within this niche, progenitors for DA neurons undergo lineage specification, migration and differentiation to become mature DA neurons (Ang, 2006; Prakash and Wurst, 2006; Smidt and Burbach, 2007; Arenas, 2008). Several lines of evidence indicate that two distinct genetic networks critically regulate the development of DA neurons. Sonic hedgehog (Shh) induces the expression of Foxa2 in vMB through specific Gli transcription factor binding elements in the enhancer sequence of Foxa2 (Sasaki et al., 1997). Interestingly, the enhancer elements in Shh contain highly conserved binding sites for Foxa2 that regulate the expression of Shh in vMB (Jeong and Epstein, 2003), supporting the notion that Shh and Foxa2 constitute a feedback transcriptional mechanism for mutual expression. Consistent with this notion, mouse mutants with region-specific removal of Foxa2 in vMB show a severe loss of Shh (Lin et al., 2009). In addition to the Shh-Foxa2 regulatory loop, the canonical Wnt/β-catenin signaling mechanism controls a distinct set of transcription factors critical for the development of DA neurons. Specifically, genetic studies in several mouse mutants indicate that Wnt1 and Otx2 form a feedback mechanism to regulate the expression for each gene (Puelles et al., 2004; Vernay et al., 2005; Prakash et al., 2006). Furthermore, in mouse embryonic stem cells (mESCs), Wnt1 and Lmx1a form a feedback regulatory mechanism similar to that in Shh-Foxa2 (Chung et al., 2009).

Several Wnts regulate the development of DA neurons in vMB. For instance, Wnt1 regulates proliferation, specification, neurogenesis in vMB DA progenitors as well as the survival of DA neurons (McMahon and Bradley, 1990; Thomas and Capecchi, 1990; Danielian and McMahon, 1996; Prakash et al., 2006). Other components of the Wnt signaling pathway, including Wnt2, the Wnt receptors Fzd3 and Fzd6, and the Wnt co-receptor, Lrp6, have been found to regulate the development of DA neurons (Sousa et al., 2009; Castelo-Branco et al., 2010; Stuebner et al., 2010). Similarly, β-catenin, a critical Wnt signaling component, is expressed in vMB DA progenitors and is required for the maintenance of adherent junctions, the integrity of radial glia processes, and cell cycle progression of DA progenitors (Joksimovic et al., 2009; Tang et al., 2009).

To further investigate the role of canonical Wnt signaling in DA neurogenesis, we generated conditional mouse mutants in which the GSK3β phosphorylation sites in β-catenin (β-CtnEx3) was removed from the neurogenic niche in vMB. Our results indicate that the activation of β-catenin in vMB promoted a marked expansion of DA progenitors, but led to a reduced expression of Shh and Foxa2. Moreover, the antagonistic interaction between the Wnt and Shh pathways in the generation of DA neurons was also detected in the cultures of DA progenitors and mESCs. Conversely, cell type specific activation of β-catenin in midline progenitors promoted DA neurogenesis. These results provide strong evidence that Wnt/β-catenin and Shh signaling pathways control a delicate balance of target gene expression during DA neurogenesis.



To generate conditional activation of β-catenin in mice, β-cateninExon3 mice (β-CtnEx3) were crossed with Shh-Cre (Jackson Lab, Stock# 005622) or Th-IRES-Cre mice (Harada et al., 1999; Harfe et al., 2004; Lindeberg et al., 2004; Tang et al., 2009). Animal care was approved by the Institutional of Animal Care and Use Committee and followed the NIH guidelines.

Histology and Immunohistochemistry (IHC)

The protocols for histology and IHC were the same as described previously (Zhang et al., 2007; Tang et al., 2009). Briefly, mouse embryos, from E10.5 to E12.5, were fixed with 1% PFA in PBS (4% PFA for Nkx6.1 antibody). Mice at embryonic day 18.5 (E18.5), postnatal day 0 (P0) and P21 were perfused and fixed with 4% PFA, and cryoprotected in 15%–30% sucrose solution and sectioned in the coronal plane using a Leica cryostat. Mouse brains were sectioned at 14 µm thickness (for stereology counting, brains were cut at 40 µm), and mounted on Superfrost glass slides.

Sections were incubated with primary antibody overnight and secondary antibodies for 1 hour, followed by incubation in DAB solution to detect signals. The primary antibodies in this study included: anti-BrdU antibody (1:500, MAB3222, Chemicon), anti-Foxa2 (1:20, 4C7, Developmental Hybridoma Study Bank [DHSB]), anti-Ki67 (1:200, RM9106-S0, Fisher), anti-Lmx1a (1:1,000, Gift of Dr. Mike German, UCSF), anti-Ngn2 (1:10, Gift of Dr. David Anderson, Caltech), anti-Pitx3 (1:300, Gift of Dr. Marten Smidt, University Medical Center Utrecht, the Netherlands), anti-Nkx2.2 (1:50, 74.5A5, DHSB), anti-Nkx6.1 (1:1,000, Gift of Dr. Mike German, UCSF), anti-Nurr1 (1:500, sc-990, Santa Cruz Biotechnology), anti-Otx2 (1:200, ab21990, Abcam), anti-phospho-histone H 3 (1:200, 06-570, Millipore), anti-Shh (1:200, #2207, Cell Signaling), anti-TuJ1 (1:2,000, MMS435P, Covance), anti-tyrosine hydroxylase (1:1,000, AB157, Chemicon), anti-tyrosine hydroxylase (1:500, ab113, Abcam), and anti-β-catenin (1:200, #9587, Cell Signaling). For stereology counting, sections were incubated for 1 hour with biotinylated IgG and avidin-biotin complex (Vector Laboratories, Burlingame, CA). Images were captured using a Nikon Eclipse E800 fluorescent microscope (Nikon, Japan), connected to a SPOT RT camera (Diagnostic Instruments Inc., Sterling Heights, MI), or a BX41 Olympus microscope equipped with Olympus DP70 CCD camera (Olympus, Japan). Images were captured using Spot Advance or Olympus DP Controller software programs or using a LSM 510 confocal microscope (Carl Zeiss 510 Microimaging).

BrdU Labeling of Dopaminergic Progenitors

We carried out 2 injection schemes. In the first scheme, the pregnant mice were injected with BrdU (50 µg/gm)(BD Bioscience) at E10.5 and E12.5, respectively, and sacrificed 2 hours later. In the second scheme, the pregnant mice were injected with BrdU at E10.5 and E11.5, respectively, and sacrificed 24 hours later (Zhang et al., 2007; Tang et al., 2009).

In Situ Hybridization

In situ hybrydization were the same as described previously (Zhang and Huang, 2006). Briefly, RNA probes for in situ hybridization were prepared using plasmid cDNA clones for Shh, cyclin D1 and Lmx1b transcribed with T7 or T3 polymerase using digoxigenin (DIG)-labeling reagents and a DIG RNA labeling kit (Roche). Embryos were fixed overnight at room temperature in 4% PFA in DEPC-treated PBS, cryoprotected in 15% and 30% sucrose in DEPC PBS, and embedded in OCT. Sections were processed at 14µm. During hybridization, sections were first postfixed with 4% PFA, washed with acetylation solution and 1% Triton X-100. Then sections were prehybridized with hybridization buffer (Amresco) for 2–4 hrs before applying hybridization buffer containing DIG-labeled riboprobes (200–400 ng/ml) at 55°C overnight. The second day, slides were washed with 4xSSC, followed by RNase A (20µg/ml) treatment at 37°C for 45min, and subsequent washes with 2xSSC, 1xSSC and 0.5xSSC at room temperature. For visualizing the in situ hybridization results, we used DIG Nucleic Acid Detection Kit (Roche). Finally, the slides were dried under room temperature and mounted with Crystal Mount (Biomeda).


The total number of TH+ neurons in SNpc and VTA was determined using the optical fractionator, an unbiased cell counting method not affected by the volume of reference (i.e. SNpc or VTA) or the size of the counted elements (i.e. neurons)(Zhang et al., 2007; Tang et al., 2009). Neuronal counts were performed using a computer-assisted image analysis system consisting of an Olympus BX-51 microscope equipped with a XYZ computer-controlled motorized stage and the StereoInvestigator software (Microbrightfield, Villiston, VT). TH+ neurons were counted in SNpc or VTA of every third section throughout the entire midbrain. Each section was viewed at lower power (4x) and outlined. At a random start, the numbers of TH-stained cells were counted at high power (60x oil; NA 1.4) using a 50×50 µm counting frame.

Ventral Midbrain (vMB) DA Progenitor Cultures

Primary cultures for dopamine neurons were prepared from ventral midbrain (vMB) using microisland methods according to published procedures (Takeshima et al., 1996). Briefly, mouse embryos were collected from time-pregnant CD-1 (for E10.5 and E13.5 wild type cultures) or Shh-Cre (for E12.5 Shh-Cre;β-CtnEx3/+ cultures) females. The ventral midbrain was dissected, dissociated after treatment with trypsin and cultured on coverslip coated with poly-D-ornithine (Sigma) and laminin (Sigma) at the density of 1.2x106/ml. The dissociated cells were maintained in the DMEM-F12 (1:1) medium containing 10% FBS overnight. Then, the differentiated neurons were changed to DMEM-F12 (1:1) medium containing N2 supplements (Gibco), 20 ng/ml FGF2 (Millipore), 100 ng/ml FGF8 (Peprotech) and designated factors, including Shh (Peprotech), Wnt1 (Peprotech), Wnt5a (R&D), and GSK3β inhibitor CT99021 (Axon Medchem) before they were fixed with 4% PFA. The number of mature DA neurons in culture were determined by counting the total number of TH+ neurons per 20x field (Parish et al., 2008).

Mouse Embryonic Stem Cell (mESC) Cultures

Differentiation of R1 mESCs into DA neurons was performed using a slightly modified protocol (Barberi et al., 2003). Briefly, R1 mESC were seeded at a density of 50 cells/cm2 on mitomycin-treated stromal cells PA6 and cultured in ES-Serum Replacement Media, composed by KnockOut-DMEM (Gibco), 15% KnockOut serum replacement (Gibco), 0.1 mM β-mercaptoethanol (Sigma), 200 mM L-Glutamine (Gibco), 1% nonessential amino acids (Biochrom AG), 2,000 U/ml penicillin/streptomycin (Gibco). After 5 days, medium was changed and supplemented with 25 ng/ml FGF8 (R&D), and different concentrations of Shh (R&D) and GSK3β inhibitor CT99021. From day 8 to day 11, cells were cultured in N2 medium consisting of F12 and MEM mixture 1:1 (Gibco), Glucose, N2 supplement (Gibco), 15 mM HEPES (Gibco), 200 mM L-Glutamine, and 3 mg/ml AlbuMax I (Gibco) supplemented with 50 ng/ml FGF8 and 10 ng/ml FGF2 (R&D), and the same concentration of Shh and CT99021 as in days 5–8. From days 11–14, the media was replaced with N2 medium supplemented with 30 ng/ml brain derived neurotrophic factor (BDNF), 30 ng/ml glial derived neurotrophic factor (GDNF, both from R&D), and 200 µM ascorbic acid (Sigma).

After in vitro differentiation, cells were fixed in 4% PFA (10’ RT), serum blocked and incubated in the appropriate primary and subsequently secondary antibodies as described previously (Parish et al., 2005). Nuclear counterstaining was performed using Hoechst. The following antibodies were used: mouse monoclonal anti-III-tubulin (TuJ1; 1:500, Promega, Madison, WI), rabbit polyclonal anti-tyrosine hydroxylase (1:500, Pel-Freez, Rogers, AK), mouse monoclonal anti-tyrosine hydroxylase (1:500, ImmunoStar), rabbit anti-Foxa2 (1:500, Cell Signaling Technology), rabbit anti-Pitx3 (1:100, Zymed Laboratories, Inc.), rabbit anti-Nurr1 (1:500, Santa Cruz Biotechnology, Inc.), Alexa Fluor 488 goat anti-mouse and Alexa Fluor 555 donkey anti-rabbit (1:500, Molecular Probes, Eugene, OR).

Statistical Analyses

Data were analyzed by two-tailed Student’s t test. Values were expressed as mean ± s.e.m. Changes were identified as significant if the p value was less than 0.05.


Activation of Wnt/β-Catenin in vMB Leads to Expansion of DA Progenitors, but Reduces DA Neurogenesis

To determine if activation of canonical Wnt/β-catenin signal in ventral midbrain (vMB) affects the development of DA neurons, we generated conditional mutant mice in which the floxed exon 3 of β-catenin (β-CtnEX3) was removed using Shh-Cre (named Shh-Cre;β-CtnEx3/+). Expression of one copy of β-CtnEX3 allele using Shh-Cre leads to perinatal lethality due to a robust gain-of-function phenotype in multiple organs, including limbs (Fig. S1 in supplementary materials). Consistent with the anticipated recombination of Shh-Cre (Tang et al., 2009), Shh-Cre;β-CtnEx3/+ mutants showed a much higher level of β-catenin protein in vMB at E12.5, with a significant accumulation of the mutant proteins in the nuclei of the neural progenitors (Fig. 1A-B and data not shown). Compared to control (β-CtnEx3/+) embryos, the vMB of Shh-Cre;β-CtnEx3/+ embryos showed a marked expansion of Sox2-, Ngn2-, and Otx2- positive progenitors in the ventricular zone (VZ)(Fig. 1C-G). In addition, DA progenitors expressing Lmx1a, Lmx1b and Nurr1 also showed significant increases in the intermediate zone (IZ) and marginal zone (MZ)(Fig. 1H-N).

Figure 1
Stabilization of β-catenin in ventral midbrain (vMB) leads to expansion of DA progenitors

We next examined whether the constitutive activation of Wnt/β-catenin in vMB could have altered cell cycle progression in DA progenitors, as previously described for Wnt1 in the neural tube (Megason and McMahon, 2002). To test this hypothesis, we performed a short-term (2 hours) BrdU labeling to determine the number of progenitors in the S-phase of cell cycle. While E10.5 and E11.5 Shh-Cre;β-CtnEx3/+ mutants showed no detectable difference in the number of BrdU+ progenitors in the vMB VZ (Fig. 2A-D), a significant increase was detected at E12.5 (Fig. 2E-F and 2M). Furthermore, a longer BrdU labeling time interval (24 hours) showed an even more drastic increase in the number of progenitors that incorporated BrdU (Fig. 2G-H & 2N). In contrast, much fewer BrdU and TH double positive neurons were generated in the Shh-Cre;β-CtnEx3/+ mutants within the same time interval (insets in Fig. 2G-H & 2O). Many of the apical progenitors in the VZ of Shh-Cre;β-CtnEx3/+ mutants continued to show positive phospho-histone H3 (PH3) staining, indicating that they were in the M phase of cell cycle (Fig. 2I-J & 2P). The increases of progenitors in S- and M-phases of cell cycle in the vMB of Shh-Cre;β-CtnEx3/+ mutants at E12.5 suggested that the constitutive activation of Wnt/β-catenin signaling may affect the cell cycle progression in DA progenitors. To address this, we performed birth-dating of DA neurons by pulse labeling the progenitors with BrdU for 24 hours and then determined the number of progenitors that have exited cell cycle (BrdU+;Ki67−) within this time interval. Consistent with our prediction, there were much fewer progenitors in the vMB of Shh-Cre;β-CtnEx3/+ mutants that have exited the cell cycle during the 24-hour time interval (Fig. 2K-L & 2Q). Together, these results supported the notion that constitutive activation of Wnt/β-catenin signal in vMB led to the expansion DA progenitors by reducing their exit from the cell cycle.

Figure 2
Cell cycle progression of the DA progenitors in vMB of Shh-Cre;β-CtnEx3/+ mutants

In analyzing the phenotype of the constitutive activation of Wnt/β-catenin signaling in DA progenitors, we noticed that the number of newly born DA neurons, marked by TH-positive staining, was reduced in the vMB of Shh-Cre;β-CtnEx3/+ mutants at E12.5 (Fig. 2E-H). To provide a more quantitative analysis of DA neurons in Shh-Cre;β-CtnEx3/+ mutants, we used stereology to determine the total number of DA neurons in vMB from E12.5 to E18.5. Our results showed that, compared to control littermates, there were consistently fewer DA neurons in the vMB of Shh-Cre;β-CtnEx3/+ mutants (Fig. 3A-G). Interestingly, a small ectopic cluster of DA neurons was identified the interpeduncular nucleus (IPN)(Fig. 3D, 3F). At E18.5, the reduction in DA neurons was more prominent in the substantia nigra pars compacta (SNpc), compared to the ventral tegmental area (VTA)(Fig. 3E-G).

Figure 3
Ventral midbrain progenitors in Shh-Cre;β-CtnEx3/+ mutants show a significant reduction in DA neurons in vivo and in vitro

To characterize the reduced DA neuron phenotype in Shh-Cre;β-CtnEx3/+ mutants, we first determined if there was an increase in cell death. Using activated caspase 3 as a marker, we found no detectable increase in cell death in the vMB of Shh-Cre;β-CtnEx3/+ mutants (Fig. S2 in supplementary materials). We next examined whether the ability of DA progenitors to differentiate was impaired in Shh-Cre;β-CtnEx3/+ mutants. To test this hypothesis, we cultured vMB progenitors from E12.5 control and Shh-Cre;β-CtnEx3/+ embryos in conditions that have previously been shown to promote differentiation of DA neurons (Takeshima et al., 1996; Ye et al., 1998; Schulte et al., 2005). Consistent with the in vivo phenotype, progenitors from Shh-Cre;β-CtnEx3/+ mutants gave rise to fewer number of DA neurons under basal culture conditions (Fig. 3H-I). Interestingly, the addition of increasing doses of Shh only promoted a very modest increase in the number of DA neurons in progenitors from Shh-Cre;β-CtnEx3/+ mutants (Fig. 3J-K & 3N). However, when treated with Wnt5a, progenitors from Shh-Cre;β-CtnEx3/+ mutant embryos showed an increase in DA neuron numbers in a manner similar to those from control (Fig. 3L-N).

Activation of Wnt/β-Catenin Antagonizes Expression of Shh and Shh Targets in vMB

The results from Figure 3 supported the notion that treatments with additional exogenous factors, such as Shh or Wnt5a, can indeed promote the generation of DA neurons from the progenitors of Shh-Cre;β-CtnEx3/+ mutants. However, the fewer number of DA neurons from Shh-Cre;β-CtnEx3/+ mutants suggested that the regional activation of canonical Wnt/β-catenin signal may have altered the milieu in the neurogenic niche of DA neurons or the intrinsic properties of DA progenitors in Shh-Cre;β-CtnEx3/+ mutants. To test these hypotheses, we examined Shh expression, an important exogenous factor that regulates the neurogenesis of DA neurons (Hynes et al., 1995). Our results showed that Shh mRNA was diffusely expressed in the floor plate at E10.5 (Fig. 4A). By E12.5, Shh mRNA became more restricted to the VZ of vMB, immediately adjacent to the neurogenic niche of DA progenitors (Fig. 4C). Despite the restricted expression pattern of Shh mRNA, Shh proteins were more widespread in the vMB, extending from VZ to the pia surface, suggesting that Shh proteins may be transported along the radial glia (Fig. 4E). This was confirmed by confocal imaging, which showed an extensive co-localization of Shh proteins with radial glia markers, e.g. Nestin, RC-2 and Glast (Fig. 4G-H and data not shown). Unlike the wild type embryos, constitutive activation of Wnt/β-catenin led to a modest decrease of Shh mRNA at E10.5 (Fig. 4B), but a near complete loss of Shh protein and mRNA in the vMB of Shh-Cre;β-CtnEx3/+ mutants at E12.5 (Fig. 4D,F). Consistent with these results, the expression of Shh targets, such as cyclin D1 and Foxa2, was reduced in the vMB of Shh-Cre;β-CtnEx3/+ mutants at E12.5, but not at E10.5 (Fig. 4I-N). In contrast, the expression of other regional vMB markers, such as Nkx2.2 and Nkx6.1, showed no detectable change (Fig. 4O-P). These results supported the hypothesis that persistent activation of Wnt/β-catenin could alter the neurogenic niche for DA neurons by antagonizing the expression of Shh and Shh target genes in the progenitors.

Figure 4
Activation of β-catenin in vMB antagonizes Shh expression and reduces Shh target genes

To further characterize the interactions between canonical Wnt/β-catenin and Shh in the generation of DA neurons, we cultured progenitors from the vMB of wild type E10.5 embryos and treated these progenitors with single, combined or sequential treatment of Shh, Wnt1 or GSK3β inhibitor CT99021 (Fig. 5A). Our results showed that treatment of these progenitors with increasing amount of recombinant Wnt1 or Shh led to a dose-dependent increase in DA neuron numbers, with the optimal concentration at 250 ng/ml (Fig. 5B-C, E-F & H-I). Consistent with these results, a selective GSK3β inhibitor CT99021 also promoted the generation of DA neurons (Fig. 5J). Surprisingly, combined treatments of Wnt1 and Shh did not show an additive or synergistic effect on the generation of DA neurons. Rather, higher doses of Wnt1 (1250 ng/ml) appeared to reduce DA neuron generation from the progenitors at the optimal condition for Shh (250 ng/ml)(Fig. 5K). Similarly, GSK3β inhibitor CT99021 also showed inhibitory effects on the generation of DA neurons in the optimal conditions for Shh (250 ng/ml)(Fig. 5L). Such antagonistic effects between Wnt1 and Shh in the generation of DA neurons were also detected in cultures obtained from the vMB of E13.5 embryos (Fig. S3 in supplementary materials).

Figure 5
Ventral midbrain progenitor cultures show antagonistic effect between canonical Wnt/β-catenin signaling and Shh

The lack of additive or synergistic effect between Wnt1 and Shh, raised the possibility that a sequential activation of canonical Wnt/β-catenin and Shh signaling pathways may be able to better recapitulate the in vivo conditions of DA neurogenesis and maximize the yield of DA neuron generation in cultures. To test this hypothesis, we cultured progenitors from E10.5 embryos and first treated them with optimal concentration of GSK3β inhibitor CT99021 or Shh (the “priming” stage) followed by switching culture conditions to optimal concentration of Shh or CT99021 (the “maintenance” stage)(Fig. 5A). Contrary to our expectations, sequential treatments with CT99021 followed by Shh, or Shh followed by CT99021, reduced the number of DA neurons compared to cultures treated with CT99021 or Shh alone (Fig. 5M-N).

The antagonistic interaction between Shh and Wnt1 in the generation of DA neurons from stem/progenitor cells was also examined in a previously established culture condition to generate DA neurons from mouse embryonic stem cells (mESCs)(Barberi et al., 2003). This culture protocol consisted of a 4-step protocol of treating mESCs co-cultured with mitomycin-treated stromal cells PA6 in serum replacement media (SRM)(Day 0 to 5), SRM plus FGF8 (Day 5 to 8), N2 media plus FGF8 and FGF2 (Day 8 to 11), and finally N2 media plus ascorbic acid, BDNF and GDNF (Day 11 to 14)(Fig. 6A). Under this condition, the majority of the TH+ neurons expressed additional dopaminergic markers, including Foxa2, Nurr1 and Pitx3a (Figure S4 in supplementary materials)(Rodriguez-Gomez et al., 2007; Hedlund et al., 2008). These results supported the notion that most TH+ neurons derived from mESCs using this protocol exhibited a phenotype consistent with that of vMB DA neurons. Our results also showed that the addition of Shh (200 ng/ml) from Day 5 to Day 11 further promoted the generation of TH+ neurons from mESCs (Fig. 6B). Unlike the primary cultures, however, addition of GSK3β inhibitor CT99021 had no effect on DA neurons (Fig. 6B). Here it is important to note that the baseline generation of DA neurons (and possibly the level of Wnt signaling) in ESC cultures is higher than in progenitors from E10.5 embryos. Despite this difference, and similar to the observation in progenitor cultures from E10.5 embryos, combined treatments of Shh and CT99021 did not show additive or synergistic effects (Fig. 6C-D). Rather, higher doses of Shh suppressed DA neurogenesis from mESCs (Fig. 6C), and high doses of CT99021 inhibited the ability of Shh to generate DA from mESCs (Fig. 6D). Moreover, we also found that high doses of CT99021 inhibited overall neurogenesis in most of the colonies, as assessed by a reduction in the number of Tuj1+ cells. Interestingly, Tuj1 positive neurons were mainly detected outside the colonies (Fig. 6D, right panel).

Figure 6
Antagonistic effects between Wnt/β-catenin and Shh in generating DA neurons from mouse embryonic stem cells (mESCs)

Activation of Wnt/β-catenin in Midline Progenitors Promotes DA Neurogenesis In Vivo

The results in Shh-Cre;β-CtnEx3/+ mutants indicated that the constitutive activation of the canonical Wnt/β-catenin signaling in the vMB led to the expansion of DA progenitors, but reduced the neurogenesis of DA neurons (Figs. 12). Based on these data, we reasoned that cell type specific activation of the Wnt/β-catenin signaling in midline progenitors may avoid the defect in DA neurogenesis seen in Shh-Cre;β-CtnEx3/+ mutants. To test this hypothesis, we generated Th-IRES-Cre;β-CtnEx3/+ mutants. We have previously shown that Th-IRES-Cre mediates recombination in essentially all postmitotic DA neurons and a subpopulation of midline progenitors at E10.5 (Tang et al., 2009). Unlike the phenotype in Th-IRES-Cre;β-Ctnfl/fl mutants, the number of DA neurons in Th-IRES-Cre;β-CtnEx3/+ mutants showed a significant increase at E11.5 and E12.5 (Fig. 7A-D and 7I). By P0 and P21, Th-IRES-Cre;β-CtnEx3/+ mutants showed ~20% increase in DA neuron numbers compared to controls (Fig. 7E-I). In addition to the increase in DA neurons, Th-IRES-Cre;β-CtnEx3/+ mutants also showed a persistent increase in the number of committed progenitors (Nurr1+;TH− cells) in vMB at E11.5 and E12.5 (Fig. 7J-N). Furthermore, we performed 24-hour neuronal birth dating experiments by labeling the progenitors with BrdU at E10.5 or E11.5, and allowed them to become TH+ postmitotic DA neurons until E11.5 and E12.5, respectively. Our results showed that the number of newly born TH+ neurons was significantly increased in Th-IRES-Cre;β-CtnEx3/+ mutants (Fig. 7O-S).

Figure 7
Increased DA neurons in Th-IRES-Cre;β-CtnEx3/+ mutants

To further investigate the mechanisms of the increased Nurr1+;TH− progenitors in Th-IRES-Cre;β-CtnEx3/+ mutants, we performed birth-dating experiments in this population by labeling the progenitors with BrdU at E10.5 or E11.5, and allowed them to develop for 24 hours. Our results showed an increase in the number of newly born Nurr1+ precursors within the 24-hour time intervals from E10.5 to E11.5, and from E11.5 to E12.5 (Fig. 8A-E). Together, these results indicated that the activation of Wnt/β-catenin signaling in a subpopulation of midline progenitors using the Th-IRES-Cre led to a significant increase in neurogenesis and DA neurons.

Figure 8
Increases in newly born postmitotic DA precursors in Th-IRES-Cre;β-CtnEx3/+ mutants


The results from this study reveal an intricate, albeit primarily antagonistic, interaction between Wnt/β-catenin and Shh during DA neurogenesis in vMB progenitors as well as in mESCs (Fig. 8F). Activation of Wnt/β-catenin can promote the expansion of DA progenitors and the generation of DA neurons. However, these effects appear to be cell-context dependent such that constitutive activation of Wnt/β-catenin in vMB using Shh-Cre expands early progenitors, but perturbs cell cycle progression in these progenitors and antagonizes the expression of Shh and Foxa2 in vMB. These phenotypes contribute to the reduced number of DA neurons. In contrast, a cell type-specific activation of Wnt/β-catenin in the midline progenitors using Th-IRES-Cre circumvents these adverse effects and leads to a significant increase in DA neuron numbers.

Wnt/β-catenin Signaling and the Development of DA Neurons

Several members of the Wnt family have been shown to regulate distinct aspects of the development of midbrain DA neurons. For instance, the canonical Wnt signaling mechanisms, mediated by Wnt1, Wnt2 and Wnt3a, control the patterning of midbrain-hindbrain junction and the initial generation of DA progenitors in vMB, whereas Wnt5a regulates the differentiation of DA neurons (Danielian and McMahon, 1996; Castelo-Branco et al., 2003; Castelo-Branco and Arenas, 2006; Andersson et al., 2008; Sousa et al., 2009). Consistent with these findings, analyses of Wnt1−/− and En1Wnt1/+ mutant mice reveal a genetic network controlled by Wnt1 to regulate the establishment of DA progenitor domain and the full differentiation of DA neurons (Prakash et al., 2006; Omodei et al., 2008). Moreover, targeted deletion of β-catenin using either region-specific Shh-Cre in vMB or cell type-specific Th-IRES-Cre in midline progenitors further demonstrate the essential role of Wnt/β-catenin signaling in the control of gene expression and in cell cycle progression during DA neurogenesis (Joksimovic et al., 2009; Tang et al., 2009). Remarkably, the effects of Wnt/β-catenin signaling appear to be highly conserved in mESCs where β-catenin and Lmx1a cooperatively controls the differentiation of DA neurons through an auto-regulatory feedback mechanism (Chung et al., 2009). Furthermore, similar roles for β-catenin have also been demonstrated in the regulation of cell cycle progression in neural progenitors of the ventral telencephalon (Gulacsi and Anderson, 2008).

Our current study provides further in vivo evidence that activation of Wnt/β-catenin signaling leads to a marked expansion of early DA progenitors that express Sox2, Ngn2 and Otx2, as well as an increase in the progenitors that express Lmx1a, Lmx1b and Nurr1 (Figs. 1 & 8F). Despite the expansion of these progenitors, however, activation of Wnt/β-catenin perturbs cell cycle progression and reduces the production of TH+ DA neurons in vMB (Figs. 2 & 3). Interestingly, when cultured in the presence of Wnt5a, the progenitors from Shh-Cre;β-CtnEx3/+ mutants differentiate into DA neurons in a manner similar to those from control (Fig. 3N). These results provide important insights into the recently published results where forced expression of Lmx1a in mESCs alone induces robust expression of Nurr1 and Pitx3, but only limited number of these cells show properties of differentiated DA neurons (Chung et al., 2009). Furthermore, our results provide additional supports that, when given the optimal growth conditions, such as excess Wnt5a, the progenitors expanded by the Wnt/β-catenin signaling mechanisms have the potential to differentiate into mature DA neurons.

Activation of Wnt/β-catenin Antagonizes Shh and Foxa2 Expression in the Neurogenesis of DA Neurons

Several explanations can account for the failure for constitutive activation of Wnt/β-catenin signaling to promote the differentiation of vMB progenitors into mature DA neurons in Shh-Cre;β-CtnEx3/+ mutants. First, as indicated above, analyses of the proliferation and cell cycle progression in the DA progenitors in Shh-Cre;β-CtnEx3/+ mutants show much more progenitors in the S- or M-phase of the cell cycle. However, these mutant progenitors show reduced cell cycle exit (Fig. 2). While the underlying cause(s) for the dysregulation of cell cycle progression in the DA progenitors of Shh-Cre;β-CtnEx3/+ mutants is not entirely clear, it is possible that the reduced expression of cyclin D1 and perhaps other cell cycle genes in the vMB of these mutants may have contributed to this phenotype. Second, the expanded progenitors may be exposed to a different environment that may prevent or delay their differentiation into committed progenitors or postmitotic neurons. Consistent with this notion, progenitors from Shh-Cre;β-CtnEx3/+ mutants can differentiate into DA neurons in the presence of Wnt5a just like those progenitors from control embryos (Fig. 3N).

The third explanation for the reduced production of DA neurons in Shh-Cre;β-CtnEx3/+ mutants is the significant down-regulation of Shh and forkhead transcription factor Foxa2 expression in the vMB (Fig. 4). The down-regulation of Shh begins as early as E10.5 and by E12.5 no detectable Shh is present in vMB in these mutants. In contrast, no detectable down-regulation of Foxa2 is present until E12.5. The down-regulation of Foxa2 may be due to the loss of Shh. Alternatively, activation of Wnt/β-catenin may directly or indirectly suppress the expression of Foxa2. Consistent with these results, expanded progenitors from Shh-Cre;β-CtnEx3/+ mutants show limited potential to differentiate into DA neurons even when cultured in the presence of excess Shh, probably due to the severe reduction in Foxa2 expression (Figs. 3N & 4L). Similar antagonistic effects of Wnt/β-catenin activation on the expression of Shh in the developing hindbrain have been reported in a recent study (Joksimovic et al., 2009). Remarkably, the antagonistic effects between Wnt/β-catenin and Shh can be demonstrated in the differentiation of DA neurons using in vitro cultures of vMB progenitors and mESCs (Figs. 5 & 6). These results support the model that Wnt/β-catenin and Shh each control distinct downstream target genes that work cooperatively to control the development of DA neurons (Fig. 8F)(Chung et al., 2009). Constitutive activation of one signaling mechanism may perturb a delicate balance between Wnt/β-catenin and Shh signaling mechanisms in the process of DA neurogenesis (Fig. 8F). Curiously, previous studies have shown that loss of Shh in the vMB of Nestin-Cre;Shhflox/flox or En1KICre/+;Shhflox/flox mutants has no detectable effects on the expression of Lmx1a, Lmx1b, Foxa1 or Foxa2 (Ferri et al., 2007; Lin et al., 2009). These studies raise the possibility that loss of Shh alone may not be sufficient to cause the phenotypes in the progenitors of Shh-Cre;β-CtnEx3/+ mutants. It is possible loss of Shh and Foxa2 in the Shh-Cre;β-CtnEx3/+ mutants cooperatively block the differentiation of DA neurons. Alternatively, activation of Wnt/β-catenin in the vMB of Shh-Cre;β-CtnEx3/+ mutants may suppress additional target genes that influence the generation of DA neurons.

The phenotype that Shh-Cre;β-CtnEx3/+ mutants show a significant reduction in Foxa2 expression in vMB is reminiscent of those in Nestin-Cre;Foxa2flox/flox mutants, which show an expansion of Nurr1+;TH− cells and a significant reduction in Nurr1+;TH+ DA neurons from E12.5 to E18.5 (Ferri et al., 2007). While Foxa1 null mutants also show a similar phenotype at E12.5, this deficit appears to be transient at E12.5 and is not detected at later developmental stages. It is unclear if the effect of β-catenin activation to suppress the expression of Foxa2 is mediated through direct binding of LEF/TCF to the enhancer sequence of Foxa2. Alternatively, it is possible that downstream targets activated by β-catenin may negatively regulate the expression of Foxa2. Regardless of the mechanism, it is most likely that down-regulation of Foxa2 may in part contributes to the reduced generation of DA neurons in these mutants.

Implications of Wnt/β-catenin and Shh in the Generation of DA Neurons from Stem Cells

The marked expansion of DA progenitors in the vMB of Shh-Cre;β-CtnEx3/+ mutants raises the possibility that Wnt/β-catenin signaling may be a feasible target to promote the generation of DA neurons from neural stem cells. In support of this, both Wnt1 and GSK3β inhibitor CT99021 promote the production of DA neurons in vMB progenitor cell and mESC cultures (Figs. 5 & 6). Despite these encouraging results, simultaneous treatment with optimal doses of Wnt1 or CT99021 and Shh in vMB progenitors or mESCs shows no additive or synergistic effects, whereas treatments with higher doses of CT99021 and Shh actually suppress the generation of DA neurons (Figs. 5 & 6). Furthermore, sequential treatments of Shh and CT99021 also do not show additional benefits (Fig. 5). These results suggest that a more detailed knowledge of the interaction between Wnt/β-catenin and Shh signaling in different vMB cells may further aid in the development of improved protocols for the generation of DA neurons in ES cell cultures. As a proof of principle, we report that cell type-specific activation of Wnt/β-catenin in midline DA progenitors, using Th-IRES-Cre, leads to increases in Nurr1+ precursors as well as in mature DA neurons both in pre- and post-natal brains (Figs. 7 & 8). Together, these encouraging results support the notion that while a broad activation of Wnt/β-catenin remains an effective means to promoting the expansion of DA progenitors, a restricted activation in midline progenitors provides beneficial effects in promoting the generation of DA neurons. We suggest that Wnt/β-catenin activation in specific cell types may become a valuable strategy to improve the DA differentiation of ES cells.

Supplementary Material



We thank Dr. David Anderson for the Ngn2 antibody, Dr. Mike German for the Lmx1a and Nkx6.1 antibodies, Dr. Marten Smidt for the Pitx3 antibody, and members of the Huang Lab for helpful discussions. This work was supported by grants from NIH (NS48393 and RR24858) and Dept of Veterans Affairs Merit Review to E.J.H., as well as grants from European Commission (Neurostemcell) and Swedish Research Council (VR2008:2811 and DBRM)(to E.A.). J.C.V. was supported by a FEBS Long-Term Fellowship and M.Z. was supported by the Development Travel Fellowship.


  • Andersson ER, Prakash N, Cajanek L, Minina E, Bryja V, Bryjova L, Yamaguchi TP, Hall AC, Wurst W, Arenas E. Wnt5a regulates ventral midbrain morphogenesis and the development of A9-A10 dopaminergic cells in vivo. PLoS ONE. 2008;3:e3517. [PMC free article] [PubMed]
  • Ang SL. Transcriptional control of midbrain dopaminergic neuron development. Development. 2006;133:3499–3506. [PubMed]
  • Arenas E. Foxa2: the rise and fall of dopamine neurons. Cell Stem Cell. 2008;2:110–112. [PubMed]
  • Barberi T, Klivenyi P, Calingasan NY, Lee H, Kawamata H, Loonam K, Perrier AL, Bruses J, Rubio ME, Topf N, Tabar V, Harrison NL, Beal MF, Moore MA, Studer L. Neural subtype specification of fertilization and nuclear transfer embryonic stem cells and application in parkinsonian mice. Nat Biotechnol. 2003;21:1200–1207. [PubMed]
  • Bjorklund A, Lindvall O. Handbook of Chemical Neuroanatomy. Amsterdam: Elsevier; 1984. Dopamine-containing systems in the CNS.
  • Castelo-Branco G, Arenas E. Function of Wnts in dopaminergic neuron development. Neurodegener Dis. 2006;3:5–11. [PubMed]
  • Castelo-Branco G, Wagner J, Rodriguez FJ, Kele J, Sousa K, Rawal N, Pasolli HA, Fuchs E, Kitajewski J, Arenas E. Differential regulation of midbrain dopaminergic neuron development by Wnt-1, Wnt-3a, and Wnt-5a. Proc Natl Acad Sci U S A. 2003;100:12747–12752. [PubMed]
  • Castelo-Branco G, Andersson ER, Minina E, Sousa KM, Ribeiro D, Kokubu C, Imai K, Prakash N, Wurst W, Arenas E. Delayed dopaminergic neuron differentiation in Lrp6 mutant mice. Dev Dyn. 2010;239:211–221. [PubMed]
  • Chung S, Leung A, Han BS, Chang MY, Moon JI, Kim CH, Hong S, Pruszak J, Isacson O, Kim KS. Wnt1-lmx1a forms a novel autoregulatory loop and controls midbrain dopaminergic differentiation synergistically with the SHH-FoxA2 pathway. Cell Stem Cell. 2009;5:646–658. [PMC free article] [PubMed]
  • Danielian PS, McMahon AP. Engrailed-1 as a target of the Wnt-1 signalling pathway in vertebrate midbrain development. Nature. 1996;383:332–334. [PubMed]
  • Ferri AL, Lin W, Mavromatakis YE, Wang JC, Sasaki H, Whitsett JA, Ang SL. Foxa1 and Foxa2 regulate multiple phases of midbrain dopaminergic neuron development in a dosage-dependent manner. Development. 2007;134:2761–2769. [PubMed]
  • Gulacsi AA, Anderson SA. Beta-catenin-mediated Wnt signaling regulates neurogenesis in the ventral telencephalon. Nat Neurosci. 2008;11:1383–1391. [PMC free article] [PubMed]
  • Harada N, Tamai Y, Ishikawa T, Sauer B, Takaku K, Oshima M, Taketo MM. Intestinal polyposis in mice with a dominant stable mutation of the beta-catenin gene. Embo J. 1999;18:5931–5942. [PubMed]
  • Harfe BD, Scherz PJ, Nissim S, Tian H, McMahon AP, Tabin CJ. Evidence for an expansion-based temporal Shh gradient in specifying vertebrate digit identities. Cell. 2004;118:517–528. [PubMed]
  • Hedlund E, Pruszak J, Lardaro T, Ludwig W, Vinuela A, Kim KS, Isacson O. Embryonic stem cell-derived Pitx3-enhanced green fluorescent protein midbrain dopamine neurons survive enrichment by fluorescence-activated cell sorting and function in an animal model of Parkinson's disease. Stem Cells. 2008;26:1526–1536. [PMC free article] [PubMed]
  • Hynes M, Porter JA, Chiang C, Chang D, Tessier-Lavigne M, Beachy PA, Rosenthal A. Induction of midbrain dopaminergic neurons by Sonic hedgehog. Neuron. 1995;15:35–44. [PubMed]
  • Jeong Y, Epstein DJ. Distinct regulators of Shh transcription in the floor plate and notochord indicate separate origins for these tissues in the mouse node. Development. 2003;130:3891–3902. [PubMed]
  • Joksimovic M, Yun BA, Kittappa R, Anderegg AM, Chang WW, Taketo MM, McKay RD, Awatramani RB. Wnt antagonism of Shh facilitates midbrain floor plate neurogenesis. Nat Neurosci. 2009;12:125–131. [PubMed]
  • Lin W, Metzakopian E, Mavromatakis YE, Gao N, Balaskas N, Sasaki H, Briscoe J, Whitsett JA, Goulding M, Kaestner KH, Ang SL. Foxa1 and Foxa2 function both upstream of and cooperatively with Lmx1a and Lmx1b in a feedforward loop promoting mesodiencephalic dopaminergic neuron development. Dev Biol. 2009;333:386–396. [PubMed]
  • Lindeberg J, Usoskin D, Bengtsson H, Gustafsson A, Kylberg A, Soderstrom S, Ebendal T. Transgenic expression of Cre recombinase from the tyrosine hydroxylase locus. Genesis. 2004;40:67–73. [PubMed]
  • McMahon AP, Bradley A. The Wnt-1 (int-1) proto-oncogene is required for development of a large region of the mouse brain. Cell. 1990;62:1073–1085. [PubMed]
  • Megason SG, McMahon AP. A mitogen gradient of dorsal midline Wnts organizes growth in the CNS. Development. 2002;129:2087–2098. [PubMed]
  • Omodei D, Acampora D, Mancuso P, Prakash N, Di Giovannantonio LG, Wurst W, Simeone A. Anterior-posterior graded response to Otx2 controls proliferation and differentiation of dopaminergic progenitors in the ventral mesencephalon. Development. 2008;135:3459–3470. [PubMed]
  • Parish CL, Parisi S, Persico MG, Arenas E, Minchiotti G. Cripto as a target for improving embryonic stem cell-based therapy in Parkinson's disease. Stem Cells. 2005;23:471–476. [PubMed]
  • Parish CL, Castelo-Branco G, Rawal N, Tonnesen J, Sorensen AT, Salto C, Kokaia M, Lindvall O, Arenas E. Wnt5a-treated midbrain neural stem cells improve dopamine cell replacement therapy in parkinsonian mice. J Clin Invest. 2008;118:149–160. [PMC free article] [PubMed]
  • Prakash N, Wurst W. Genetic networks controlling the development of midbrain dopaminergic neurons. J Physiol. 2006;575:403–410. [PubMed]
  • Prakash N, Brodski C, Naserke T, Puelles E, Gogoi R, Hall A, Panhuysen M, Echevarria D, Sussel L, Weisenhorn DM, Martinez S, Arenas E, Simeone A, Wurst W. A Wnt1-regulated genetic network controls the identity and fate of midbrain-dopaminergic progenitors in vivo. Development. 2006;133:89–98. [PubMed]
  • Puelles E, Annino A, Tuorto F, Usiello A, Acampora D, Czerny T, Brodski C, Ang SL, Wurst W, Simeone A. Otx2 regulates the extent, identity and fate of neuronal progenitor domains in the ventral midbrain. Development. 2004;131:2037–2048. [PubMed]
  • Rodriguez-Gomez JA, Lu JQ, Velasco I, Rivera S, Zoghbi SS, Liow JS, Musachio JL, Chin FT, Toyama H, Seidel J, Green MV, Thanos PK, Ichise M, Pike VW, Innis RB, McKay RD. Persistent dopamine functions of neurons derived from embryonic stem cells in a rodent model of Parkinson disease. Stem Cells. 2007;25:918–928. [PubMed]
  • Sasaki H, Hui C, Nakafuku M, Kondoh H. A binding site for Gli proteins is essential for HNF-3beta floor plate enhancer activity in transgenics and can respond to Shh in vitro. Development. 1997;124:1313–1322. [PubMed]
  • Schulte G, Bryja V, Rawal N, Castelo-Branco G, Sousa KM, Arenas E. Purified Wnt-5a increases differentiation of midbrain dopaminergic cells and dishevelled phosphorylation. J Neurochem. 2005;92:1550–1553. [PubMed]
  • Smidt MP, Burbach JP. How to make a mesodiencephalic dopaminergic neuron. Nat Rev Neurosci. 2007;8:21–32. [PubMed]
  • Sousa KM, Villaescusa JC, Cajanek L, Ondr JK, Castelo-Branco G, Hofstra W, Bryja V, Palmberg C, Bergman T, Wainwright B, Lang RA, Arenas E. WNT2 regulates progenitor proliferation in the developing ventral midbrain. J Biol Chem. 2009 [PubMed]
  • Stuebner S, Faus-Kessler T, Fischer T, Wurst W, Prakash N. Fzd3 and Fzd6 deficiency results in a severe midbrain morphogenesis defect. Dev Dyn. 2010;239:246–260. [PubMed]
  • Takeshima T, Shimoda K, Johnston JM, Commissiong JW. Standardized methods to bioassay neurotrophic factors for dopaminergic neurons. J Neurosci Methods. 1996;67:27–41. [PubMed]
  • Tang M, Miyamoto Y, Huang EJ. Multiple roles of beta-catenin in controlling the neurogenic niche for midbrain dopamine neurons. Development. 2009;136:2027–2038. [PubMed]
  • Thomas KR, Capecchi MR. Targeted disruption of the murine int-1 proto-oncogene resulting in severe abnormalities in midbrain and cerebellar development. Nature. 1990;346:847–850. [PubMed]
  • Vernay B, Koch M, Vaccarino F, Briscoe J, Simeone A, Kageyama R, Ang SL. Otx2 regulates subtype specification and neurogenesis in the midbrain. J Neurosci. 2005;25:4856–4867. [PubMed]
  • Ye W, Shimamura K, Rubenstein JL, Hynes MA, Rosenthal A. FGF and Shh signals control dopaminergic and serotonergic cell fate in the anterior neural plate. Cell. 1998;93:755–766. [PubMed]
  • Zhang J, Huang EJ. Dynamic expression of neurotrophic factor receptors in postnatal spinal motoneurons and in mouse model of ALS. J Neurobiol. 2006;66:882–895. [PMC free article] [PubMed]
  • Zhang J, Pho V, Bonasera SJ, Holzmann J, Tang AT, Hellmuth J, Tang S, Janak PH, Tecott LH, Huang EJ. Essential function of HIPK2 in TGFbeta-dependent survival of midbrain dopamine neurons. Nat Neurosci. 2007;10:77–86. [PMC free article] [PubMed]