Cell Reports
Report Sox6 and Otx2 Control the Specification of Substantia Nigra and Ventral Tegmental Area Dopamine Neurons Lia Panman,1,2,* Maria Papathanou,1 Ariadna Laguna,1,6 Tony Oosterveen,2 Nikolaos Volakakis,1 Dario Acampora,3,4 Idha Kurtsdotter,1,6 Takashi Yoshitake,5 Jan Kehr,5 Eliza Joodmardi,1 Jonas Muhr,1,6 Antonio Simeone,3,4 Johan Ericson,6 and Thomas Perlmann1,6,* 1Ludwig
Institute for Cancer Research, 17177 Stockholm, Sweden Toxicology Unit, Leicester LE1 9HN, UK 3Institute of Genetics and Biophysics ‘‘A. Buzzati-Traverso,’’ CNR, 80131 Naples, Italy 4IRCCS Neuromed, Pozzilli IS 86077, Italy 5Department of Physiology and Pharmacology, Karolinska Institutet, 17177 Stockholm, Sweden 6Department of Cell and Molecular Biology, Karolinska Institutet, 17177 Stockholm, Sweden *Correspondence: [email protected] (L.P.), [email protected] (T.P.) http://dx.doi.org/10.1016/j.celrep.2014.07.016 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). 2MRC
SUMMARY
Distinct midbrain dopamine (mDA) neuron subtypes are found in the substantia nigra pars compacta (SNc) and the ventral tegmental area (VTA), but it is mainly SNc neurons that degenerate in Parkinson’s disease. Interest in how mDA neurons develop has been stimulated by the potential use of stem cells in therapy or disease modeling. However, very little is known about how specific dopaminergic subtypes are generated. Here, we show that the expression profiles of the transcription factors Sox6, Otx2, and Nolz1 define subpopulations of mDA neurons already at the neural progenitor cell stage. After cell-cycle exit, Sox6 selectively localizes to SNc neurons, while Otx2 and Nolz1 are expressed in a subset of VTA neurons. Importantly, Sox6 ablation leads to decreased expression of SNc markers and a corresponding increase in VTA markers, while Otx2 ablation has the opposite effect. Moreover, deletion of Sox6 affects striatal innervation and dopamine levels. We also find reduced Sox6 levels in Parkinson’s disease patients. These findings identify Sox6 as a determinant of SNc neuron development and should facilitate the engineering of relevant mDA neurons for cell therapy and disease modeling. INTRODUCTION Midbrain dopamine (mDA) neurons constitute a highly diverse neuronal population controlling important brain functions such as voluntary movements, cognition, and reward. Progressive degeneration of mDA neurons causes severe motor symptoms in Parkinson’s disease, and altered dopamine (DA) neurotrans-
mission is associated with addiction and psychiatric disease. mDA neurons can be subdivided into major anatomically and functionally distinct cell groups, including the substantia nigra pars compacta (SNc) (A9) and ventral tegmental area (VTA; A10) (Bjo¨rklund and Dunnett, 2007; McRitchie et al., 1996). SNc neurons innervate the dorsal/lateral striatum and are critical for controlling motor functions, while neurons within the VTA innervate cortical and limbic areas and are primarily associated with emotional behaviors. Notably, cells within the SNc are the most vulnerable mDA neurons and are severely affected in Parkinson’s disease (Damier et al., 1999; Fearnley and Lees, 1991). All mDA neurons are derived from a common population of neural progenitor cells localized within the ventral midbrain and caudal diencephalon (Marı´n et al., 2005). Signaling molecules including Sonic hedgehog (Shh), fibroblast growth factor (FGF) 8, and members of the Wnt family of secreted glycoproteins play key roles in establishing expression of transcription factors such as Lmx1a, Lmx1b, Otx2, and FoxA2 in mDA neural progenitors (e.g., Andersson et al., 2006; Castelo-Branco et al., 2003; Ferri et al., 2007; Prakash et al., 2006; Puelles et al., 2004; Smidt et al., 2000). Proliferating progenitor cells will eventually exit the cell cycle, migrate toward their final destinations, and begin to express pan-neuronal and mDA neuron-specific genes, including tyrosine hydroxylase (TH) and the dopamine transporter (DAT). Although considerable progress has been made in understanding key events determining the generation of all mDA neurons, remarkably little is known of how distinct subpopulations are specified during embryogenesis. Previous characterization has revealed molecular differences between SNc and VTA neurons, but it has remained unclear at what developmental stage SNc and VTA neurons first become molecularly distinct (Chung et al., 2005; Wolfart et al., 2001). Recently, it was shown that the transcription factor Otx2 is involved in controlling the subtype identity of VTA neurons. Otx2 is first expressed in neural progenitor cells and maintained in a subset of differentiating and mature VTA neurons (Chung et al., 2010; Di Salvio et al., 2010b). Gain- and loss-of-function analysis both in progenitor cells and postmitotic neurons demonstrated a role of Otx2 in
1018 Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors
P120
E18.5 Sox6TH
Sox6TH
Nolz1TH
Nolz1TH
B
C
D
E
F
G
H
VTA
SNc
A
FG
I
FGTH
J
FGTH
FGSox6
K
L
SN
SN
VTA
SN
Figure 1. Sox6 and Nolz1 Are Selectively Expressed in SNc and VTA Dopamine Neurons, Respectively (A–H) Immunohistochemical analysis of Sox6 and Nolz1 expression in combination with TH in SNc and VTA neurons in E18.5 embryos and P120 adult mice. Arrowheads indicate Sox6-expressing mDA neurons in SNc (A and C) and the few expressing cells in the VTA (E and G). Arrowheads in (F) and (H) point to Nolz1-expressing cells in VTA dopamine neurons. (I–L) Fluorogold (FG) retrograde tracing analysis. (I) FG was unilaterally injected in the adult dorsal-lateral striatum (n = 4). The injection site in the striatum (outlined by dashed line) is evident 24 hr after injection. (J) FG selectively labels SNc dopamine neurons, but not the VTA. (K) High-magnification image of SNc showing FG labeling of TH+ neurons. (L) Most FG-positive SNc dopamine neurons are positive for Sox6. Scale bars, 100 mm. See also Figure S1.
promoting VTA subtype identity (Chung et al., 2010; Di Giovannantonio et al., 2013; Di Salvio et al., 2010a). In addition, SNc neurons are selectively lost in mice deficient for the transcription factor Pitx3 (Hwang et al., 2003; Nunes et al., 2003; Smidt et al., 2004; van den Munckhof et al., 2003). However, Pitx3 is expressed at a relatively late postmitotic stage in all mDA neurons and is not involved in early specification events. Therefore, additional transcription factors controlling lineage selection into SNc mDA neurons need to be identified. We previously identified the transcription factors Sox6 and Nolz1 (also referred to as Zfp503) in a screen for mDA neuron-expressed genes (Panman et al., 2011). Both Sox6 and Nolz1 are important for neuronal subtype diversification in other brain regions (Azim et al., 2009; Batista-Brito et al., 2009; Chang et al., 2013; Ji et al., 2009). Here, we show that the expression of Sox6 and Nolz1 define distinct populations of mDA neuron progenitors and identify Sox6 as a key transcription factor for the specification and development of SNc mDA neurons. RESULTS Sox6, Otx2, and Nolz1 Are Selectively Expressed in Distinct Dopaminergic Neuronal Subpopulations We examined the localization of Sox6 and Nolz1 in embryonic and adult dopaminergic neurons. Immunohistochemistry
showed Sox6 or Nolz1 expression in distinct mDA neuron subpopulations at embryonic day 18.5 (E18.5) and in the adult brain (Figure 1; Figure S1). While Sox6 was expressed in the lateral population of TH-positive neurons corresponding to the SNc (approximately 95% at E18.5 and 85% in adults) and in only a few scattered cells in the dorsal VTA, Nolz1 was selectively expressed in a subpopulation of VTA neurons in the E18.5 (approximately 18%; data not shown) and adult (postnatal day 120 [P120]; fewer than 10%) ventral midbrain (Figures 1A–1H). Analyses of additional markers confirmed the restricted expression of Sox6 to SNc neurons that innervate the dorsal striatum. Sox6 expression in the SNc overlaps with Girk2 and glycosylated DA transporter (GlycoDAT) and partly with Raldh1, but not with the VTA markers Calbindin and Otx2 at E18.5 and in adults (Figure S1). We performed retrograde tracing by injecting fluorogold into the adult dorsal-lateral striatum leading to retrograde labeling of neurons within the SNc, but not within the VTA (Figures 1I– 1K). Importantly, the majority of fluorogold-positive cells were also positive for Sox6 (Figure 1L). Moreover, Sox6 expression was diminished while Nolz1 was maintained in Pitx3 mutant mice, which is consistent with selective loss of SNc mDA neurons in these animals (Figure S2). Thus, Sox6 expression is predominantly restricted to mDA neurons that innervate the dorsal-lateral striatum. Sox6 is also expressed in interneurons in the dorsal midbrain at E15.5 and E18.5 as shown by in situ hybridization and in additional uncharacterized cell types (Figure S2 and data not shown). The expression of Sox6, Nolz1, and Otx2 was analyzed at stages when mDA neurons become specified (E11.5 and E13.5; Figure 2). The transcription factor Lmx1a was selectively expressed in all progenitor cells that give rise to mDA neurons (Andersson et al., 2006). In contrast, Sox6, Nolz1, and Otx2 were expressed in distinct subpopulations of mDA neuron progenitor cells at E11.5 (Figures 2A–2E). While Sox6 was detected in the medial Corin+ and Lmx1a+ progenitor domain, Nolz1 expression was localized laterally (Figures 2A–2E; Figure S2). Double Sox6 and Nolz1 staining showed that the expression of these factors was mutually exclusive in medial and lateral mDA progenitors, respectively (Figure 2D). In addition, Otx2 was expressed only at low levels in medial Sox6+ progenitors (Figures 2C and 2E) but was strongly expressed within Nolz1+ lateral cells (Figures 2C and 2E). Thus, mDA neuron progenitors are organized in molecularly distinct medial (Sox6+/Otx2weak/Nolz1) and lateral (Sox6/Otx2strong/Nolz1+) domains. In early postmitotic neurons, a few cells express both Sox6 and Otx2 (arrows in Figures 2A, 2C, and 2D). These cells appear to be immature mDA precursor cells that express the general mDA neuron transcription factor Nurr1, but not TH (Figure S2). To further understand how SNc and VTA neurons are specified, expression of Sox6, Otx2, and Nolz1 was analyzed after progenitor cell-cycle exit when cells are acquiring dopaminergic traits and begin to migrate toward their final destinations (Figures 2F–2Q). At E13.5, Sox6 and Nolz1/Otx2 were expressed in a nonoverlapping pattern in postmitotic differentiating neurons (Figures 2F–2K). Postmitotic Sox6+ mDA neurons were found medially within the intermediate zone in the mDA domain and extended laterally along the marginal zone (Figures 2F, 2G, and 2I). Thus, cells leaving the medial Sox6+ progenitor zone seemed
Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors 1019
B
E11.5
*
E13.5
Sox6Nolz1
F
Sox6Nolz1 L
Sox6Otx2
Nolz1Lmx1a
Sox6Lmx1a A
C
Sox6Otx2
Sox6Otx2 M
E
*
*
*
G
Otx2Nolz1
Sox6Nolz1 D
Otx2Nolz1
H
I
Sox6TH J
Otx2Nolz1 R
*
Nolz1TH
Otx2TH
E13.5
E11.5
N
K
pZ mDA
E18.5
O
P
Q
SN
VTA
E18.5
SNc
pZ
mDA
VTA
Sox6+Lmx1a+ Otx2+Lmx1a+ Nolz1+Otx2+Lmx1a+ Lmx1a-
Figure 2. Spatial and Temporal Dynamics of Sox6, Nolz1, and Otx2 Expression in the Developing Ventral Midbrain (A–E) Combinatorial expression analysis using antibodies against Sox6, Lmx1a, Nolz1, and Otx2 in E11.5 ventral midbrain reveals a distinct lateral and medial progenitor domain. Medially located Sox6 expression within Lmx1a+ progenitor domain (asterisk). Arrows in (A) and (D) point to immature postmitotic mDA neuron precursors. Arrows in (C) point to Otx2+Sox6+ postmitotic mDA precursor cells. (F–K) Expression of Sox6, Nolz1, Otx2, and TH at E13.5. A subset of Otx2+ neurons expresses Nolz1 (arrows in H), while the more medially positioned Otx2+ cells are Nolz1 (arrowheads). Sox6 is expressed in medially located TH+, while its expression extends laterally at the marginal zone (arrows in I). Nolz1- and Otx2expressing TH-positive neurons are located laterally (arrows in J and K). (L–Q) Expression of Sox6, Nolz1, and Otx2 in the SNc and VTA at E18.5. Sox6 is expressed in SNc dopamine neurons (L and M), while Otx2 and Nolz1 are absent (L, M, and N). (O–Q) Nolz1 and Otx2 are expressed in VTA dopamine neurons. Few Sox6-positive mDA neurons can be found in the dorsal-lateral VTA (arrowheads in O and P). Arrows in (P) point to rarely observed Sox6+Otx2+ neurons. (Q) Nolz1 is expressed in a subset of Otx2+ VTA neurons (arrowheads). (R) Schematic representation summarizing the distribution of cell populations at E11.5, E13.5, and E18.5. See text for details. Scale bars, 100 mm. See also Figure S2.
to migrate ventrally and then laterally to end up in the lateral SNc at E18.5 (Figures 2L–2Q). Postmitotic Otx2+ and Nolz1+ cells were instead positioned in two vertical stripes extending ventrally from the lateral Nolz1+ progenitor domain (Figures 2F, 2G, 2J, and 2K). At E13.5, Nolz1 was already confined to the most lateral subpopulation of Otx2+ neurons in a pattern that was maintained in the VTA at E18.5, where Nolz1 was also expressed in the most dorsal-lateral Otx2+ VTA neurons (Figures 2F, 2H, 2J, and 2Q). These findings lead to two conclusions concerning the origin of SNc and VTA neurons (illustrated in Figure 2R). First, proliferating cells within the Lmx1a+ progenitor zone have already acquired a distinct molecular character that is related to the subsequent generation of SNc and VTA neurons. This is defined by the mutually exclusive expression of Sox6 or Nolz1 and the differential expression of Otx2 (weak in the medial but strong in the lateral region). Second, the expression pattern indicates a dynamic ‘‘inside-out’’ migration of medial Sox6+/Otx2weak cells to the positions occupied by laterally located SNc-forming mDA neurons, while Nolz1+/Otx2strong cells originate from a
lateral position in the ventricular progenitor zone to settle in the medially located VTA (Figure 2R). Otx2 Suppression of Sox6 and Altered SNc-Specific Neuronal Character in Sox6/ Mutant Embryos The complementary expression of Sox6 and Otx2 during early mDA neuron progenitor specification (medial: Sox6+/Otx2low; lateral: Sox6/Otx2high) indicates that these transcription factors might cross-regulate each other. Indeed, by conditional ablation of Otx2 in ventral progenitors (Di Giovannantonio et al., 2013), we note that Sox6 expression was expanded laterally (Figure 3A). As shown previously, Otx2 ablation results in a severely diminished generation of VTA neurons (Di Giovannantonio et al., 2013). Interestingly, remaining postmitotic neurons positioned in the VTA region of Otx2-ablated embryos expressed ectopic Sox6 (Figure 3A). In addition, overexpression of Otx2 in postmitotic SNc neurons in vivo suppressed Sox6 expression (Figure S3), while ablation of Otx2 in postmitotic neurons did not result in the expansion of Sox6 expression domain. Thus, Otx2 is essential for restricting the expression of Sox6 in medial mDA progenitors
1020 Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors
En1Cre;Otx2fl/fl
*
*
*
Otx2TH
*
Figure 3. Otx2 and Sox6 Have Distinct Roles in mDA Subtype Specification
Sox6-/-
Wild-type
B
Sox6TH
Sox6
Wild-type
Otx2
A
VTA
VTA
SN
SN
CalbTH Raldh1TH GlycoDatTH E C D
F Wild-type Sox6-/-
%X/TH+
Wild-type
40
** **
20
*
Sox6-/-
Calb+ Raldh1+GlycoDat-
G
Wild-type
Sox6-/-
**
H
Wild-type Sox6-/-
I
Wild-type
1 2
Epha5
DAT-DAB
1 2
Relative fiber density
100%
area 1 area 2
Sox6-/-
(A) Ablation of Otx2 in E11.5 ventral midbrain results in a lateral expansion of the Sox6 expression domain (indicated with a bracket). Asterisk indicates the ventral midline. (B) Otx2 expression remains unaltered in ventral midbrain of E12.5 Sox6 mutant embryos. (C) Immunohistochemistry showing ventral midbrain distribution of Calbindin (Calb) and TH in the SNc at E18.5 in wild-type and Sox6/ embryos, respectively. Arrows point to Calb-labeled TH+ neurons. (D) Immunohistochemistry showing ventral midbrain distribution of Raldh1 and TH in the SNc at E18.5 in wild-type and Sox6/ embryos, respectively. Arrows point to Raldh1-labeled TH+ neurons. (E) Immunohistochemistry showing ventral midbrain distribution of GlycoDAT and TH in the SNc at 18.5 in wild-type and Sox6/ embryos, respectively. Arrows point to GlycoDat-negative TH+ neurons. (F) Percentage of TH+ neurons expressing Calbindin (Calb), GlycoDat, or Raldh1 in wild-type and Sox6/ embryos (n = 3), respectively; mean values ± SD; **p < 0.01. (G) Images represent coronal sections through the striatum of an E18.5 embryo stained with Dat to visualize Dat-expressing fibers. Area 1 marks the lateral part of the striatum, while area 2 marks the medial part. (H) Immunohistochemistry of DAT innervation analyzed by densitometry in a lateral (area1) and medial (area 2) region of the striatum in wild-type and Sox6/ embryos, respectively (n = 4); mean values ± SD; *p < 0.05, **p < 0.01. (I) In situ hybridization using a probe against Epha5 in wild-type and Sox6/ embryos, respectively. Dashed circle marks the location of SNc dopamine neurons in the E18.5 embryo. Scale bars, 100 mm. See also Figure S3.
DAT-DAB
and sufficient for the suppression of Sox6 expression in postmitotic mDA neurons. To analyze the role of Sox6 in suppression of Otx2 and in the generation of SNc mDA neurons, Sox6 null (Sox6/) mutant embryos were studied (Dumitriu et al., 2006). Expression of neither Otx2 nor Nolz1 was increased in Sox6/ mutant progenitor cells or postmitotic neurons (Figure 3B and data not shown). However, although the number of TH+ dopaminergic neurons was unchanged (data not shown), notable differences between genotypes were seen. The number of neurons within the SNc that expressed the VTA marker Calbindin was markedly increased in E18.5 Sox6/ embryos (Figures 3C and 3F). The glycosylated form of DAT (GlycoDAT) was strongly expressed in SNc mDA neurons and in a dorsal-lateral population in the VTA in wildtype embryos. In contrast, neurons that were negative for GlycoDAT expression were significantly increased in Sox6/ SNc (Figures 3E and 3F), while expression of the Raldh1 in the SNc was reduced (Figures 3D and 3F).
In addition, at E18.5, striatal fiber innervation was affected in Sox6/ mutant embryos as shown by 3,30 -diaminobenzidine immunoperoxidase (DAB) staining for DAT and TH, respectively (Figures 3G and 3H; Figure S4). Notably, the lateral striatal region (area 1; Figure 3G), which is mainly innervated by SNc mDA neurons, was markedly less intensely stained as shown by densitometry using antibodies against both DAT and TH (Figure 3H; Figure S4). The more medial (area 2) and ventral striatum, which is predominantly innervated by VTA mDA neurons, showed normal TH and DAT immunostaining intensity (Figures 3G and 3H; Figure S4). In addition, expression of the ephrin receptor Epha5 was selectively expressed in SNc neurons in wild-type embryos and strongly reduced in Sox6/ animals, as determined by in situ hybridization analysis (Figure 3I). Thus, the reduced innervation of the lateral part of the striatum could be due to a defect in the axonal guidance pathway, but it may also be influenced by other deficiencies in SNc mDA neurons. The expression of several other axon-guidance molecules and
Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors 1021
general mDA neuron markers including Lmx1a, Pitx3, En1, and Nurr1 were unchanged (data not shown). Taken together, these data show that Otx2 plays a critical role in the suppression of Sox6 in mDA neuron progenitors and postmitotic neurons. Moreover, although a complete transition in phenotype from SNc to VTA neurons does not occur as a consequence of Sox6 ablation, the results clearly indicate that Sox6 plays an important role in promoting SNc-specific and suppressing VTA-specific characteristics in mDA neurons. Sox6 Is Essential in Postmitotic SNc Neurons and Is Expressed in Human SNc mDA Neurons Sox6/ mice do not survive after birth because of deficiencies that are unrelated to the phenotype in mDA neurons (Smits et al., 2001). To analyze the role of Sox6 in postmitotic mDA neurons, floxed Sox6 mice were crossed with DatCre mice to selectively inactivate Sox6 in Dat-expressing postmitotic mDA neurons (Kadkhodaei et al., 2009). Within the ventral midbrain, the number of mDA neurons appeared intact in DatCre/+;Sox6fl/fl mutant embryos and adult mice (data not shown). However, + there was a reduced density of dendritic TH fibers in the ventral midbrain of 4- and 8-month-old mutant mice (Figure 4D). In addition, while the striatal innervation in E18.5 DatCre/+;Sox6fl/fl mutant embryos appeared normal (Figure S3), TH+ fiber density in the striatum of 4-month-old mutant mice was clearly reduced (area 1; Figures 4A and 4C; Figure S4). The intensity of mDA neuron fiber density was further decreased in the dorsal-lateral part of the striatum of DatCre/+;Sox6fl/fl in 8-month-old animals (Figures 4B and 4C; Figure S4), while the innervation of the ventral striatum remained unaffected (Figures 4A–4C). Importantly, the affected striatal region is primarily innervated from Sox6+ SNc mDA neurons. Other analyzed regions within the hippocampus and amygdala appeared unaffected (data not shown). In addition, TH+ cell bodies within the VTA in 8-monthold mutant mice also appeared normal (Figure S4). Furthermore, consistent with a decreased striatal fiber innervation, DA and DA metabolites (DOPAC and HVA) were reduced in the caudate putamen (CPu) of Sox6 conditionally ablated mice (Figure 4E), while levels within VTA-innervated nucleus accumbens were not significantly altered (Figure S4). Further, the open field test showed that Sox6 conditional knockout mice were more active compared to controls (Figure S4), a behavior that has previously been shown to result from prenatal SNc mDA neuron deficiency (e.g., Hwang et al., 2005). Thus, ablation of Sox6 within differentiating mDA neurons demonstrates an important role in maintaining SNc, but not VTA, characteristics in mDA neurons. The results showing that Sox6 is important for maintaining normal postmitotic SNc mDA neuron properties in mice raise the question of how Sox6 is expressed in the normal and diseased human brain. Postmortem tissue samples from both Parkinson’s disease brains and aged-matched controls were analyzed by Sox6 immunostaining (Figures 4F–4H). Sox6 staining was clearly detectable and localized to neuromelanin and TH-positive neurons in the SNc (Figure 4F; Figure S4). Sox6 expression was not observed in medial neuromelanin-negative neurons (data not shown), but we could not firmly conclude whether these neurons are VTA neurons. Thus, Sox6 expression in the human mDA neurons is likely recapitulating that seen in
mice. To analyze whether Sox6 expression is affected in Parkinson’s disease, we compared the levels of Sox6 expression by densitometry in 145 individual neuromelanin-positive cells in controls (n = 4) and 323 neuromelanin-positive cells in Parkinson’s disease tissue samples (n = 8). Interestingly, a significant reduction of Sox6 staining in Parkinson’s disease patients compared to controls was evident when considering measurements in all individual neurons (Figure 4G) and averages in staining intensities determined for each control and each Parkinson’s disease patient (Figure 4H). Thus, Sox6 expression is clearly diminished in neuromelanin-positive neurons in Parkinson’s disease patients. DISCUSSION The importance of mDA neurons in normal brain function and their degeneration in Parkinson’s disease has placed this cell population at central stage in studies of lineage-specific neurogenesis during CNS development. An important motivation to understand how these cells develop originates from strong interest in developing methods for stem cell engineering of mDA neurons to be used in cell therapy or for disease modeling in cell culture (Cooper et al., 2012; Lindvall and Bjo¨rklund, 2011; Studer, 2012). However, although signaling and transcription factor pathways controlling generic aspects of mDA neuron specification and differentiation are beginning to be understood, the processes regulating mDA neuron diversity have remained mostly unclear. In this study, we elucidated a transcription factor code that subdivides the Lmx1a+ progenitor pool in lateral (Otx2highNolz1+Sox6) and medial (Otx2lowNolz1Sox6+) domains, respectively. After cell-cycle exit, these transcription factors are expressed in a corresponding way in VTA (Otx2+Nolz1+Sox6) and SNc (OtxNolz1Sox6+) postmitotic mDA neurons. This coordinated expression before and after cell-cycle exit suggests that subtype identities have already been determined in mDA neural progenitors. The finding that Sox6+ progenitor cells reside in medial domain while Nolz1+ and Otx2high progenitors are found in a lateral Sox6 region also indicated an intriguing, and somewhat counterintuitive, inside-out migration of developing SNc mDA neurons (Figure 2R). Interestingly, recent fate-mapping studies using temporally controlled Cre expression in reporter mice support the conclusion that SNc neurons originate from medial mDA progenitor cells (Blaess et al., 2011; Bodea et al., 2014; Joksimovic et al., 2009). Using two different mouse ablation strategies, we demonstrated that Sox6 is required for the normal development of SNc neurons and maintenance of some SNc-specific neuronal properties. Together, the data reveal both early and late roles for Sox6 in SNc neurons. Our finding that Sox6 ablation by DatCre leads to decreased density of TH+ fibers in the dorsal-lateral striatum and a corresponding decrease in the levels of striatal DA is intriguing when considering our observation that Sox6 expression is diminished in neuromelanin-positive neurons in Parkinson’s disease patients. Interestingly, a recent study identified a mutation within the human Sox6 gene leading to a general developmental delay and parkinsonian symptoms in an affected child (Scott et al., 2014). A decrease in Sox6 in association with
1022 Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors
B
4 months DatCre;Sox6
fl/+
DatCre;Sox6
1
fl/fl
*
**
1
2
2
2
area 1
area 2 area 1
TH-DAB
DatCre;Sox6fl/fl
DatCre;Sox6fl/+
DatCre;Sox6fl/+
area 2
DatCre;Sox6fl/fl
TH-DAB
D
Control DatCre;Sox6fl/fl
100%
1
TH-DAB
4 months 8 months
DatCre;Sox6fl/fl
DatCre;Sox6fl/+
1
2
C
8 months
Relative fiber density
A
**
p=0.052
*
Control DatCre;Sox6fl/fl
Sox6neuromelanin
F
Relative levels
100%
DA
H
DOPAC HVA
CPu Control 2
150 100 50 0
PD 1
PD 2
Control
PD
150
Densitometry (AU)
Control 1
***
200
G Densitometry (AU)
E
100
*
50
0
Control
PD
Figure 4. Sox6 Is Required for the Functional Integrity of SNc Neurons in Adult Mice (A) Immunohistochemical analysis of TH+ fiber innervation in the striatum of 4-month-old DatCre;Sox6fl/+ and DatCre;Sox6fl/fl mice. Area 1 represents the dorsallateral part of the striatum, while area 2 represents the ventral part of the striatum. (B) Immunohistochemical analysis of TH expression in the striatum of 8-month-old DatCre;Sox6fl/+ and DatCre;Sox6fl/fl mice. (C) Analysis of TH-expressing fibers in the striatum by densitometry in a dorsolateral (area1) and medial (area 2) region of the striatum at 4 and 8 months in DatCre;Sox6fl/+ and DatCre;Sox6fl/fl mice, respectively. (n = 4); mean values ± SD; *p < 0.05; **p < 0.01. (D) Analysis of TH expression in midbrain of 4- and 8-month-old DatCre;Sox6fl/+ and DatCre;Sox6fl/fl mice. Scale bars in (A)–(D), 200 mm. (E) DA and DA metabolites (DOPAC and HVA) analyzed by high-performance liquid chromatography from extracts of the dorsolateral part of the striatum (area 1) of 6-month-old DatCre;Sox6fl/+ and DatCre;Sox6fl/fl mice (n = 10). Values are shown as percentages relative the levels detected in extracts from DatCre;Sox6fl/+ mice (100%); mean values ± SD; *p < 0.05; **p < 0.01. (F) Sox6 expression in the human SNc. Expression analyzed by immunohistochemistry for Sox6 is shown in sections from two controls and two individuals with Parkinson’s disease (PD1 and PD2). Distribution of Sox6 (dark blue) is localized to neuromelanin-positive (brown) neurons in the SNc of control individuals (arrows), while nuclear Sox6 expression is not observed in PD patient 1 and only weakly observed in some nuclei of PD patient 2 (arrowheads). Scale bar, 25 mm. (G) Intensity of Sox6 levels in individual neuromelanin-positive neurons (n = 323) is decreased in in Parkinson’s disease patients compared to neurons (n = 145) in controls. Intensity was measured by densitometry and shown for each cell in arbitrary units; mean values ± SEM; ***p < 0.001. (H) The intensities from the individual cells (shown in G) is averaged for each Parkinson’s disease patient (n = 8) or control (n = 4) and determined in arbitrary units; mean values ± SEM; *p < 0.05. See also Figure S4.
Parkinson’s disease pathology may thus contribute to continuing deterioration of SNc mDA neurons in this disease. From the data reported here and from previous studies on Otx2, it should be noted that neither Otx2 nor Sox6 ablation led to a complete cell-fate transition of subtype identities (Di Giovan-
nantonio et al., 2013; Di Salvio et al., 2010b). Thus, we postulate that additional factors contribute to lineage selection of VTA and SNc neurons. It is interesting that Nolz1 is expressed strictly in lateral neural progenitors in which Otx2 is also highly expressed. An intriguing possibility is that Otx2 and Nolz1 functionally
Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors 1023
interact to determine the VTA neuronal fate. It is also important to note that the diversity of mDA neuron types extends beyond the division into SNc and VTA neurons. Thus, another major mDA neuron group localized in the retrorubral field (A8) has properties that are distinct from those in the SNc and VTA (Bjo¨rklund and Dunnett, 2007), and within the substantia nigra and VTA, molecularly and functionally diverse mDA neuron subpopulations have also been described (see e.g. Di Salvio et al., 2010a; Fu et al., 2011; Roeper, 2013). How Sox6, Otx2, Nolz1, and other transcription factors influence additional complexity within mDA neurons therefore remains an important question to investigate. The studies presented here have important implications for stem cell engineering. Cell transplantation in Parkinson’s disease animal models has demonstrated that Girk2+ SNc neurons must be a major constituent of cell grafts to restore lost motor function in Parkinson’s disease (Cooper et al., 2012; Grealish et al., 2010; Zuddas et al., 1991). The identification of Sox6 and Nolz1 as two early progenitor markers for prospective SNc and VTA neurons, respectively, will provide a valuable tool in tracing how different culturing conditions affect lineage selection at very early stages in stem cell differentiation experiments. This, in turn, should facilitate possibilities to identify culturing conditions resulting in SNc or VTA mDA neuron-enriched cultures (Studer, 2012). EXPERIMENTAL PROCEDURES Mouse Strains Sox6fl/+ animals (Dumitriu et al., 2006) were crossed with DatCre/+;Sox6fl/+ animals (Ekstrand et al., 2007) to obtain DatCre/+;Sox6fl/fl mutant offspring. Sox6/ mutant embryos were obtained from crossings between Sox6+/ heterozygous mice. Otx2 mutant mouse strains were kept and bred as described previously (Di Giovannantonio et al., 2013; Di Salvio et al., 2010a). All animal experiments were approved by the local animal ethics committee and conform to the relevant regulatory standards. For details, see Supplemental Experimental Procedures. Immunohistochemistry Embryos were fixed for 1–3 hr in 4% paraformaldehyde (PFA) and further processed as described before (Briscoe et al., 2000). Adult mice were perfused with 4% PFA as described previously (Kadkhodaei et al., 2009), and dissected brains were postfixed for 4 hr in 4%PFA. For details, see Supplemental Experimental Procedures. Human Postmortem Tissue Paraffin-embedded midbrain sections from Parkinson’s disease patients (n = 8) and age-matched control individuals (n = 4) were provided by the UK Parkinson’s Disease Society Tissue Bank at Imperial College. For details, see Supplemental Experimental Procedures. Statistical Analysis Statistical significance was calculated by the one-paired Student’s t test, and data are presented as mean ± SD (*p < 0.05, **p < 0.01). SUPPLEMENTAL INFORMATION
experiment; A.L. analyzed human tissue; T.O. performed histology; D.A. and A.S. analyzed Otx2 ablated embryos; N.V. helped write the manuscript; I.K. bred mice; T.Y. and J.K. performed high-performance liquid chromatography analysis; E.J. performed histology; J.M. and J.E. helped with analysis and planning; and T.P. together with L.P. planned all experiments and wrote the manuscript. ACKNOWLEDGMENTS We thank Bhumica Singla and Hilda Lunden-Miguel for technical assistance. We thank Veronique Lefebvre and Meng Li for generously providing mouse lines. This work was supported by funding from the European Union, Seventh Framework Programme under grant agreement mdDANeurodev and NeuroStemCell (to T.P.), from the Swedish Strategic Research Foundation (SSF; to T.P.), and from Vetenskapsra˚det (L.P. and T.P.) and Parkinsonfonden (L.P.) Received: December 27, 2013 Revised: June 24, 2014 Accepted: July 14, 2014 Published: August 7, 2014 REFERENCES Andersson, E., Tryggvason, U., Deng, Q., Friling, S., Alekseenko, Z., Robert, B., Perlmann, T., and Ericson, J. (2006). Identification of intrinsic determinants of midbrain dopamine neurons. Cell 124, 393–405. Azim, E., Jabaudon, D., Fame, R.M., and Macklis, J.D. (2009). SOX6 controls dorsal progenitor identity and interneuron diversity during neocortical development. Nat. Neurosci. 12, 1238–1247. Batista-Brito, R., Rossignol, E., Hjerling-Leffler, J., Denaxa, M., Wegner, M., Lefebvre, V., Pachnis, V., and Fishell, G. (2009). The cell-intrinsic requirement of Sox6 for cortical interneuron development. Neuron 63, 466–481. Bjo¨rklund, A., and Dunnett, S.B. (2007). Dopamine neuron systems in the brain: an update. Trends Neurosci. 30, 194–202. Blaess, S., Bodea, G.O., Kabanova, A., Chanet, S., Mugniery, E., Derouiche, A., Stephen, D., and Joyner, A.L. (2011). Temporal-spatial changes in Sonic Hedgehog expression and signaling reveal different potentials of ventral mesencephalic progenitors to populate distinct ventral midbrain nuclei. Neural Dev. 6, 29. Bodea, G.O., Spille, J.H., Abe, P., Andersson, A.S., Acker-Palmer, A., Stumm, R., Kubitscheck, U., and Blaess, S. (2014). Reelin and CXCL12 regulate distinct migratory behaviors during the development of the dopaminergic system. Development 141, 661–673. Briscoe, J., Pierani, A., Jessell, T.M., and Ericson, J. (2000). A homeodomain protein code specifies progenitor cell identity and neuronal fate in the ventral neural tube. Cell 101, 435–445. Castelo-Branco, G., Wagner, J., Rodriguez, F.J., Kele, J., Sousa, K., Rawal, N., Pasolli, H.A., Fuchs, E., Kitajewski, J., and Arenas, E. (2003). Differential regulation of midbrain dopaminergic neuron development by Wnt-1, Wnt-3a, and Wnt-5a. Proc. Natl. Acad. Sci. USA 100, 12747–12752. Chang, S.L.-Y., Chen, S.-Y., Huang, H.-H., Ko, H.-A., Liu, P.-T., Liu, Y.-C., Chen, P.-H., and Liu, F.-C. (2013). Ectopic expression of nolz-1 in neural progenitors promotes cell cycle exit/premature neuronal differentiation accompanying with abnormal apoptosis in the developing mouse telencephalon. PLoS ONE 8, e74975.
Supplemental Information includes Supplemental Experimental Procedures and four figures and can be found with this article online at http://dx.doi.org/ 10.1016/j.celrep.2014.07.016.
Chung, C.Y., Seo, H., Sonntag, K.C., Brooks, A., Lin, L., and Isacson, O. (2005). Cell type-specific gene expression of midbrain dopaminergic neurons reveals molecules involved in their vulnerability and protection. Hum. Mol. Genet. 14, 1709–1725.
AUTHOR CONTRIBUTIONS
Chung, C.Y., Licznerski, P., Alavian, K.N., Simeone, A., Lin, Z., Martin, E., Vance, J., and Isacson, O. (2010). The transcription factor orthodenticle homeobox 2 influences axonal projections and vulnerability of midbrain dopaminergic neurons. Brain 133, 2022–2031.
L.P. planned all experiments, performed histological analysis, and wrote the manuscript; M.P. performed behavioral analysis and the fluorogold tracing
1024 Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors
Cooper, O., Parmar, M., and Isacson, O. (2012). Characterization and criteria of embryonic stem and induced pluripotent stem cells for a dopamine replacement therapy. Prog. Brain Res. 200, 265–276. Damier, P., Hirsch, E.C., Agid, Y., and Graybiel, A.M. (1999). The substantia nigra of the human brain. II. Patterns of loss of dopamine-containing neurons in Parkinson’s disease. Brain 122, 1437–1448. Di Giovannantonio, L.G., Di Salvio, M., Acampora, D., Prakash, N., Wurst, W., and Simeone, A. (2013). Otx2 selectively controls the neurogenesis of specific neuronal subtypes of the ventral tegmental area and compensates En1dependent neuronal loss and MPTP vulnerability. Dev. Biol. 373, 176–183.
Nurr1 is required for maintenance of maturing and adult midbrain dopamine neurons. J. Neurosci. 29, 15923–15932. Lindvall, O., and Bjo¨rklund, A. (2011). Cell therapeutics in Parkinson’s disease. Neurotherapeutics 8, 539–548. Marı´n, F., Herrero, M.-T., Vyas, S., and Puelles, L. (2005). Ontogeny of tyrosine hydroxylase mRNA expression in mid- and forebrain: neuromeric pattern and novel positive regions. Dev. Dyn. 234, 709–717. McRitchie, D.A., Hardman, C.D., and Halliday, G.M. (1996). Cytoarchitectural distribution of calcium binding proteins in midbrain dopaminergic regions of rats and humans. J. Comp. Neurol. 364, 121–150.
Di Salvio, M., Di Giovannantonio, L.G., Omodei, D., Acampora, D., and Simeone, A. (2010a). Otx2 expression is restricted to dopaminergic neurons of the ventral tegmental area in the adult brain. Int. J. Dev. Biol. 54, 939–945.
Nunes, I., Tovmasian, L.T., Silva, R.M., Burke, R.E., and Goff, S.P. (2003). Pitx3 is required for development of substantia nigra dopaminergic neurons. Proc. Natl. Acad. Sci. USA 100, 4245–4250.
Di Salvio, M., Di Giovannantonio, L.G., Acampora, D., Prosperi, R., Omodei, D., Prakash, N., Wurst, W., and Simeone, A. (2010b). Otx2 controls neuron subtype identity in ventral tegmental area and antagonizes vulnerability to MPTP. Nat. Neurosci. 13, 1481–1488.
Panman, L., Andersson, E., Alekseenko, Z., Hedlund, E., Kee, N., Mong, J., Uhde, C.W., Deng, Q., Sandberg, R., Stanton, L.W., et al. (2011). Transcription factor-induced lineage selection of stem-cell-derived neural progenitor cells. Cell Stem Cell 8, 663–675.
Dumitriu, B., Dy, P., Smits, P., and Lefebvre, V. (2006). Generation of mice harboring a Sox6 conditional null allele. Genesis 44, 219–224.
Prakash, N., Brodski, C., Naserke, T., Puelles, E., Gogoi, R., Hall, A., Panhuysen, M., Echevarria, D., Sussel, L., Weisenhorn, D.M.V., et al. (2006). A Wnt1-regulated genetic network controls the identity and fate of midbraindopaminergic progenitors in vivo. Development 133, 89–98.
Ekstrand, M.I., Terzioglu, M., Galter, D., Zhu, S., Hofstetter, C., Lindqvist, E., Thams, S., Bergstrand, A., Hansson, F.S., Trifunovic, A., et al. (2007). Progressive parkinsonism in mice with respiratory-chain-deficient dopamine neurons. Proc. Natl. Acad. Sci. USA 104, 1325–1330. Fearnley, J.M., and Lees, A.J. (1991). Ageing and Parkinson’s disease: substantia nigra regional selectivity. Brain 114, 2283–2301. Ferri, A.L.M., Lin, W., Mavromatakis, Y.E., Wang, J.C., Sasaki, H., Whitsett, J.A., and Ang, S.-L. (2007). Foxa1 and Foxa2 regulate multiple phases of midbrain dopaminergic neuron development in a dosage-dependent manner. Development 134, 2761–2769. Fu, Y., Yuan, Y., Halliday, G., Ruszna´k, Z., Watson, C., and Paxinos, G. (2011). A cytoarchitectonic and chemoarchitectonic analysis of the dopamine cell groups in the substantia nigra, ventral tegmental area, and retrorubral field in the mouse. Brain Struct. Funct. 217, 591–612. Grealish, S., Jo¨nsson, M.E., Li, M., Kirik, D., Bjo¨rklund, A., and Thompson, L.H. (2010). The A9 dopamine neuron component in grafts of ventral mesencephalon is an important determinant for recovery of motor function in a rat model of Parkinson’s disease. Brain 133, 482–495. Hwang, D.-Y., Ardayfio, P., Kang, U.J., Semina, E.V., and Kim, K.-S. (2003). Selective loss of dopaminergic neurons in the substantia nigra of Pitx3-deficient aphakia mice. Brain Res. Mol. Brain Res. 114, 123–131. Hwang, D.-Y., Fleming, S.M., Ardayfio, P., Moran-Gates, T., Kim, H., Tarazi, F.I., Chesselet, M.-F., and Kim, K.-S. (2005). 3,4-dihydroxyphenylalanine reverses the motor deficits in Pitx3-deficient aphakia mice: behavioral characterization of a novel genetic model of Parkinson’s disease. J. Neurosci. 25, 2132–2137. Ji, S.-J., Periz, G., and Sockanathan, S. (2009). Nolz1 is induced by retinoid signals and controls motoneuron subtype identity through distinct repressor activities. Development 136, 231–240. Joksimovic, M., Anderegg, A., Roy, A., Campochiaro, L., Yun, B., Kittappa, R., McKay, R., and Awatramani, R. (2009). Spatiotemporally separable Shh domains in the midbrain define distinct dopaminergic progenitor pools. Proc. Natl. Acad. Sci. USA 106, 19185–19190. Kadkhodaei, B., Ito, T., Joodmardi, E., Mattsson, B., Rouillard, C., Carta, M., Muramatsu, S., Sumi-Ichinose, C., Nomura, T., Metzger, D., et al. (2009).
Puelles, E., Annino, A., Tuorto, F., Usiello, A., Acampora, D., Czerny, T., Brodski, C., Ang, S.-L., Wurst, W., and Simeone, A. (2004). Otx2 regulates the extent, identity and fate of neuronal progenitor domains in the ventral midbrain. Development 131, 2037–2048. Roeper, J. (2013). Dissecting the diversity of midbrain dopamine neurons. Trends Neurosci. 36, 336–342. Scott, O., Pugh, J., Kiddoo, D., Sonnenberg, L.K., Bamforth, S., and Goez, H.R. (2014). Global developmental delay, progressive relapsing-remitting parkinsonism, and spinal syrinx in a child with SOX6 mutation. J. Child Neurol., Published online January 21, 2014 http://dx.doi.org/10.1177/0883073813514134. Smidt, M.P., Asbreuk, C.H., Cox, J.J., Chen, H., Johnson, R.L., and Burbach, J.P. (2000). A second independent pathway for development of mesencephalic dopaminergic neurons requires Lmx1b. Nat. Neurosci. 3, 337–341. Smidt, M.P., Smits, S.M., Bouwmeester, H., Hamers, F.P.T., van der Linden, A.J.A., Hellemons, A.J.C.G.M., Graw, J., and Burbach, J.P.H. (2004). Early developmental failure of substantia nigra dopamine neurons in mice lacking the homeodomain gene Pitx3. Development 131, 1145–1155. Smits, P., Li, P., Mandel, J., Zhang, Z., Deng, J.M., Behringer, R.R., de Crombrugghe, B., and Lefebvre, V. (2001). The transcription factors L-Sox5 and Sox6 are essential for cartilage formation. Dev. Cell 1, 277–290. Studer, L. (2012). Derivation of dopaminergic neurons from pluripotent stem cells. Prog. Brain Res. 200, 243–263. van den Munckhof, P., Luk, K.C., Ste-Marie, L., Montgomery, J., Blanchet, P.J., Sadikot, A.F., and Drouin, J. (2003). Pitx3 is required for motor activity and for survival of a subset of midbrain dopaminergic neurons. Development 130, 2535–2542. Wolfart, J., Neuhoff, H., Franz, O., and Roeper, J. (2001). Differential expression of the small-conductance, calcium-activated potassium channel SK3 is critical for pacemaker control in dopaminergic midbrain neurons. J. Neurosci. 21, 3443–3456. Zuddas, A., Corsini, G.U., Barker, J.L., Kopin, I.J., and Di Porzio, U. (1991). Specific Reinnervation of Lesioned Mouse Striatum by Grafted Mesencephalic Dopaminergic Neurons. Eur. J. Neurosci. 3, 72–85.
Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors 1025
Report Sox6 and Otx2 Control the Specification of Substantia Nigra and Ventral Tegmental Area Dopamine Neurons Lia Panman,1,2,* Maria Papathanou,1 Ariadna Laguna,1,6 Tony Oosterveen,2 Nikolaos Volakakis,1 Dario Acampora,3,4 Idha Kurtsdotter,1,6 Takashi Yoshitake,5 Jan Kehr,5 Eliza Joodmardi,1 Jonas Muhr,1,6 Antonio Simeone,3,4 Johan Ericson,6 and Thomas Perlmann1,6,* 1Ludwig
Institute for Cancer Research, 17177 Stockholm, Sweden Toxicology Unit, Leicester LE1 9HN, UK 3Institute of Genetics and Biophysics ‘‘A. Buzzati-Traverso,’’ CNR, 80131 Naples, Italy 4IRCCS Neuromed, Pozzilli IS 86077, Italy 5Department of Physiology and Pharmacology, Karolinska Institutet, 17177 Stockholm, Sweden 6Department of Cell and Molecular Biology, Karolinska Institutet, 17177 Stockholm, Sweden *Correspondence: [email protected] (L.P.), [email protected] (T.P.) http://dx.doi.org/10.1016/j.celrep.2014.07.016 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). 2MRC
SUMMARY
Distinct midbrain dopamine (mDA) neuron subtypes are found in the substantia nigra pars compacta (SNc) and the ventral tegmental area (VTA), but it is mainly SNc neurons that degenerate in Parkinson’s disease. Interest in how mDA neurons develop has been stimulated by the potential use of stem cells in therapy or disease modeling. However, very little is known about how specific dopaminergic subtypes are generated. Here, we show that the expression profiles of the transcription factors Sox6, Otx2, and Nolz1 define subpopulations of mDA neurons already at the neural progenitor cell stage. After cell-cycle exit, Sox6 selectively localizes to SNc neurons, while Otx2 and Nolz1 are expressed in a subset of VTA neurons. Importantly, Sox6 ablation leads to decreased expression of SNc markers and a corresponding increase in VTA markers, while Otx2 ablation has the opposite effect. Moreover, deletion of Sox6 affects striatal innervation and dopamine levels. We also find reduced Sox6 levels in Parkinson’s disease patients. These findings identify Sox6 as a determinant of SNc neuron development and should facilitate the engineering of relevant mDA neurons for cell therapy and disease modeling. INTRODUCTION Midbrain dopamine (mDA) neurons constitute a highly diverse neuronal population controlling important brain functions such as voluntary movements, cognition, and reward. Progressive degeneration of mDA neurons causes severe motor symptoms in Parkinson’s disease, and altered dopamine (DA) neurotrans-
mission is associated with addiction and psychiatric disease. mDA neurons can be subdivided into major anatomically and functionally distinct cell groups, including the substantia nigra pars compacta (SNc) (A9) and ventral tegmental area (VTA; A10) (Bjo¨rklund and Dunnett, 2007; McRitchie et al., 1996). SNc neurons innervate the dorsal/lateral striatum and are critical for controlling motor functions, while neurons within the VTA innervate cortical and limbic areas and are primarily associated with emotional behaviors. Notably, cells within the SNc are the most vulnerable mDA neurons and are severely affected in Parkinson’s disease (Damier et al., 1999; Fearnley and Lees, 1991). All mDA neurons are derived from a common population of neural progenitor cells localized within the ventral midbrain and caudal diencephalon (Marı´n et al., 2005). Signaling molecules including Sonic hedgehog (Shh), fibroblast growth factor (FGF) 8, and members of the Wnt family of secreted glycoproteins play key roles in establishing expression of transcription factors such as Lmx1a, Lmx1b, Otx2, and FoxA2 in mDA neural progenitors (e.g., Andersson et al., 2006; Castelo-Branco et al., 2003; Ferri et al., 2007; Prakash et al., 2006; Puelles et al., 2004; Smidt et al., 2000). Proliferating progenitor cells will eventually exit the cell cycle, migrate toward their final destinations, and begin to express pan-neuronal and mDA neuron-specific genes, including tyrosine hydroxylase (TH) and the dopamine transporter (DAT). Although considerable progress has been made in understanding key events determining the generation of all mDA neurons, remarkably little is known of how distinct subpopulations are specified during embryogenesis. Previous characterization has revealed molecular differences between SNc and VTA neurons, but it has remained unclear at what developmental stage SNc and VTA neurons first become molecularly distinct (Chung et al., 2005; Wolfart et al., 2001). Recently, it was shown that the transcription factor Otx2 is involved in controlling the subtype identity of VTA neurons. Otx2 is first expressed in neural progenitor cells and maintained in a subset of differentiating and mature VTA neurons (Chung et al., 2010; Di Salvio et al., 2010b). Gain- and loss-of-function analysis both in progenitor cells and postmitotic neurons demonstrated a role of Otx2 in
1018 Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors
P120
E18.5 Sox6TH
Sox6TH
Nolz1TH
Nolz1TH
B
C
D
E
F
G
H
VTA
SNc
A
FG
I
FGTH
J
FGTH
FGSox6
K
L
SN
SN
VTA
SN
Figure 1. Sox6 and Nolz1 Are Selectively Expressed in SNc and VTA Dopamine Neurons, Respectively (A–H) Immunohistochemical analysis of Sox6 and Nolz1 expression in combination with TH in SNc and VTA neurons in E18.5 embryos and P120 adult mice. Arrowheads indicate Sox6-expressing mDA neurons in SNc (A and C) and the few expressing cells in the VTA (E and G). Arrowheads in (F) and (H) point to Nolz1-expressing cells in VTA dopamine neurons. (I–L) Fluorogold (FG) retrograde tracing analysis. (I) FG was unilaterally injected in the adult dorsal-lateral striatum (n = 4). The injection site in the striatum (outlined by dashed line) is evident 24 hr after injection. (J) FG selectively labels SNc dopamine neurons, but not the VTA. (K) High-magnification image of SNc showing FG labeling of TH+ neurons. (L) Most FG-positive SNc dopamine neurons are positive for Sox6. Scale bars, 100 mm. See also Figure S1.
promoting VTA subtype identity (Chung et al., 2010; Di Giovannantonio et al., 2013; Di Salvio et al., 2010a). In addition, SNc neurons are selectively lost in mice deficient for the transcription factor Pitx3 (Hwang et al., 2003; Nunes et al., 2003; Smidt et al., 2004; van den Munckhof et al., 2003). However, Pitx3 is expressed at a relatively late postmitotic stage in all mDA neurons and is not involved in early specification events. Therefore, additional transcription factors controlling lineage selection into SNc mDA neurons need to be identified. We previously identified the transcription factors Sox6 and Nolz1 (also referred to as Zfp503) in a screen for mDA neuron-expressed genes (Panman et al., 2011). Both Sox6 and Nolz1 are important for neuronal subtype diversification in other brain regions (Azim et al., 2009; Batista-Brito et al., 2009; Chang et al., 2013; Ji et al., 2009). Here, we show that the expression of Sox6 and Nolz1 define distinct populations of mDA neuron progenitors and identify Sox6 as a key transcription factor for the specification and development of SNc mDA neurons. RESULTS Sox6, Otx2, and Nolz1 Are Selectively Expressed in Distinct Dopaminergic Neuronal Subpopulations We examined the localization of Sox6 and Nolz1 in embryonic and adult dopaminergic neurons. Immunohistochemistry
showed Sox6 or Nolz1 expression in distinct mDA neuron subpopulations at embryonic day 18.5 (E18.5) and in the adult brain (Figure 1; Figure S1). While Sox6 was expressed in the lateral population of TH-positive neurons corresponding to the SNc (approximately 95% at E18.5 and 85% in adults) and in only a few scattered cells in the dorsal VTA, Nolz1 was selectively expressed in a subpopulation of VTA neurons in the E18.5 (approximately 18%; data not shown) and adult (postnatal day 120 [P120]; fewer than 10%) ventral midbrain (Figures 1A–1H). Analyses of additional markers confirmed the restricted expression of Sox6 to SNc neurons that innervate the dorsal striatum. Sox6 expression in the SNc overlaps with Girk2 and glycosylated DA transporter (GlycoDAT) and partly with Raldh1, but not with the VTA markers Calbindin and Otx2 at E18.5 and in adults (Figure S1). We performed retrograde tracing by injecting fluorogold into the adult dorsal-lateral striatum leading to retrograde labeling of neurons within the SNc, but not within the VTA (Figures 1I– 1K). Importantly, the majority of fluorogold-positive cells were also positive for Sox6 (Figure 1L). Moreover, Sox6 expression was diminished while Nolz1 was maintained in Pitx3 mutant mice, which is consistent with selective loss of SNc mDA neurons in these animals (Figure S2). Thus, Sox6 expression is predominantly restricted to mDA neurons that innervate the dorsal-lateral striatum. Sox6 is also expressed in interneurons in the dorsal midbrain at E15.5 and E18.5 as shown by in situ hybridization and in additional uncharacterized cell types (Figure S2 and data not shown). The expression of Sox6, Nolz1, and Otx2 was analyzed at stages when mDA neurons become specified (E11.5 and E13.5; Figure 2). The transcription factor Lmx1a was selectively expressed in all progenitor cells that give rise to mDA neurons (Andersson et al., 2006). In contrast, Sox6, Nolz1, and Otx2 were expressed in distinct subpopulations of mDA neuron progenitor cells at E11.5 (Figures 2A–2E). While Sox6 was detected in the medial Corin+ and Lmx1a+ progenitor domain, Nolz1 expression was localized laterally (Figures 2A–2E; Figure S2). Double Sox6 and Nolz1 staining showed that the expression of these factors was mutually exclusive in medial and lateral mDA progenitors, respectively (Figure 2D). In addition, Otx2 was expressed only at low levels in medial Sox6+ progenitors (Figures 2C and 2E) but was strongly expressed within Nolz1+ lateral cells (Figures 2C and 2E). Thus, mDA neuron progenitors are organized in molecularly distinct medial (Sox6+/Otx2weak/Nolz1) and lateral (Sox6/Otx2strong/Nolz1+) domains. In early postmitotic neurons, a few cells express both Sox6 and Otx2 (arrows in Figures 2A, 2C, and 2D). These cells appear to be immature mDA precursor cells that express the general mDA neuron transcription factor Nurr1, but not TH (Figure S2). To further understand how SNc and VTA neurons are specified, expression of Sox6, Otx2, and Nolz1 was analyzed after progenitor cell-cycle exit when cells are acquiring dopaminergic traits and begin to migrate toward their final destinations (Figures 2F–2Q). At E13.5, Sox6 and Nolz1/Otx2 were expressed in a nonoverlapping pattern in postmitotic differentiating neurons (Figures 2F–2K). Postmitotic Sox6+ mDA neurons were found medially within the intermediate zone in the mDA domain and extended laterally along the marginal zone (Figures 2F, 2G, and 2I). Thus, cells leaving the medial Sox6+ progenitor zone seemed
Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors 1019
B
E11.5
*
E13.5
Sox6Nolz1
F
Sox6Nolz1 L
Sox6Otx2
Nolz1Lmx1a
Sox6Lmx1a A
C
Sox6Otx2
Sox6Otx2 M
E
*
*
*
G
Otx2Nolz1
Sox6Nolz1 D
Otx2Nolz1
H
I
Sox6TH J
Otx2Nolz1 R
*
Nolz1TH
Otx2TH
E13.5
E11.5
N
K
pZ mDA
E18.5
O
P
Q
SN
VTA
E18.5
SNc
pZ
mDA
VTA
Sox6+Lmx1a+ Otx2+Lmx1a+ Nolz1+Otx2+Lmx1a+ Lmx1a-
Figure 2. Spatial and Temporal Dynamics of Sox6, Nolz1, and Otx2 Expression in the Developing Ventral Midbrain (A–E) Combinatorial expression analysis using antibodies against Sox6, Lmx1a, Nolz1, and Otx2 in E11.5 ventral midbrain reveals a distinct lateral and medial progenitor domain. Medially located Sox6 expression within Lmx1a+ progenitor domain (asterisk). Arrows in (A) and (D) point to immature postmitotic mDA neuron precursors. Arrows in (C) point to Otx2+Sox6+ postmitotic mDA precursor cells. (F–K) Expression of Sox6, Nolz1, Otx2, and TH at E13.5. A subset of Otx2+ neurons expresses Nolz1 (arrows in H), while the more medially positioned Otx2+ cells are Nolz1 (arrowheads). Sox6 is expressed in medially located TH+, while its expression extends laterally at the marginal zone (arrows in I). Nolz1- and Otx2expressing TH-positive neurons are located laterally (arrows in J and K). (L–Q) Expression of Sox6, Nolz1, and Otx2 in the SNc and VTA at E18.5. Sox6 is expressed in SNc dopamine neurons (L and M), while Otx2 and Nolz1 are absent (L, M, and N). (O–Q) Nolz1 and Otx2 are expressed in VTA dopamine neurons. Few Sox6-positive mDA neurons can be found in the dorsal-lateral VTA (arrowheads in O and P). Arrows in (P) point to rarely observed Sox6+Otx2+ neurons. (Q) Nolz1 is expressed in a subset of Otx2+ VTA neurons (arrowheads). (R) Schematic representation summarizing the distribution of cell populations at E11.5, E13.5, and E18.5. See text for details. Scale bars, 100 mm. See also Figure S2.
to migrate ventrally and then laterally to end up in the lateral SNc at E18.5 (Figures 2L–2Q). Postmitotic Otx2+ and Nolz1+ cells were instead positioned in two vertical stripes extending ventrally from the lateral Nolz1+ progenitor domain (Figures 2F, 2G, 2J, and 2K). At E13.5, Nolz1 was already confined to the most lateral subpopulation of Otx2+ neurons in a pattern that was maintained in the VTA at E18.5, where Nolz1 was also expressed in the most dorsal-lateral Otx2+ VTA neurons (Figures 2F, 2H, 2J, and 2Q). These findings lead to two conclusions concerning the origin of SNc and VTA neurons (illustrated in Figure 2R). First, proliferating cells within the Lmx1a+ progenitor zone have already acquired a distinct molecular character that is related to the subsequent generation of SNc and VTA neurons. This is defined by the mutually exclusive expression of Sox6 or Nolz1 and the differential expression of Otx2 (weak in the medial but strong in the lateral region). Second, the expression pattern indicates a dynamic ‘‘inside-out’’ migration of medial Sox6+/Otx2weak cells to the positions occupied by laterally located SNc-forming mDA neurons, while Nolz1+/Otx2strong cells originate from a
lateral position in the ventricular progenitor zone to settle in the medially located VTA (Figure 2R). Otx2 Suppression of Sox6 and Altered SNc-Specific Neuronal Character in Sox6/ Mutant Embryos The complementary expression of Sox6 and Otx2 during early mDA neuron progenitor specification (medial: Sox6+/Otx2low; lateral: Sox6/Otx2high) indicates that these transcription factors might cross-regulate each other. Indeed, by conditional ablation of Otx2 in ventral progenitors (Di Giovannantonio et al., 2013), we note that Sox6 expression was expanded laterally (Figure 3A). As shown previously, Otx2 ablation results in a severely diminished generation of VTA neurons (Di Giovannantonio et al., 2013). Interestingly, remaining postmitotic neurons positioned in the VTA region of Otx2-ablated embryos expressed ectopic Sox6 (Figure 3A). In addition, overexpression of Otx2 in postmitotic SNc neurons in vivo suppressed Sox6 expression (Figure S3), while ablation of Otx2 in postmitotic neurons did not result in the expansion of Sox6 expression domain. Thus, Otx2 is essential for restricting the expression of Sox6 in medial mDA progenitors
1020 Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors
En1Cre;Otx2fl/fl
*
*
*
Otx2TH
*
Figure 3. Otx2 and Sox6 Have Distinct Roles in mDA Subtype Specification
Sox6-/-
Wild-type
B
Sox6TH
Sox6
Wild-type
Otx2
A
VTA
VTA
SN
SN
CalbTH Raldh1TH GlycoDatTH E C D
F Wild-type Sox6-/-
%X/TH+
Wild-type
40
** **
20
*
Sox6-/-
Calb+ Raldh1+GlycoDat-
G
Wild-type
Sox6-/-
**
H
Wild-type Sox6-/-
I
Wild-type
1 2
Epha5
DAT-DAB
1 2
Relative fiber density
100%
area 1 area 2
Sox6-/-
(A) Ablation of Otx2 in E11.5 ventral midbrain results in a lateral expansion of the Sox6 expression domain (indicated with a bracket). Asterisk indicates the ventral midline. (B) Otx2 expression remains unaltered in ventral midbrain of E12.5 Sox6 mutant embryos. (C) Immunohistochemistry showing ventral midbrain distribution of Calbindin (Calb) and TH in the SNc at E18.5 in wild-type and Sox6/ embryos, respectively. Arrows point to Calb-labeled TH+ neurons. (D) Immunohistochemistry showing ventral midbrain distribution of Raldh1 and TH in the SNc at E18.5 in wild-type and Sox6/ embryos, respectively. Arrows point to Raldh1-labeled TH+ neurons. (E) Immunohistochemistry showing ventral midbrain distribution of GlycoDAT and TH in the SNc at 18.5 in wild-type and Sox6/ embryos, respectively. Arrows point to GlycoDat-negative TH+ neurons. (F) Percentage of TH+ neurons expressing Calbindin (Calb), GlycoDat, or Raldh1 in wild-type and Sox6/ embryos (n = 3), respectively; mean values ± SD; **p < 0.01. (G) Images represent coronal sections through the striatum of an E18.5 embryo stained with Dat to visualize Dat-expressing fibers. Area 1 marks the lateral part of the striatum, while area 2 marks the medial part. (H) Immunohistochemistry of DAT innervation analyzed by densitometry in a lateral (area1) and medial (area 2) region of the striatum in wild-type and Sox6/ embryos, respectively (n = 4); mean values ± SD; *p < 0.05, **p < 0.01. (I) In situ hybridization using a probe against Epha5 in wild-type and Sox6/ embryos, respectively. Dashed circle marks the location of SNc dopamine neurons in the E18.5 embryo. Scale bars, 100 mm. See also Figure S3.
DAT-DAB
and sufficient for the suppression of Sox6 expression in postmitotic mDA neurons. To analyze the role of Sox6 in suppression of Otx2 and in the generation of SNc mDA neurons, Sox6 null (Sox6/) mutant embryos were studied (Dumitriu et al., 2006). Expression of neither Otx2 nor Nolz1 was increased in Sox6/ mutant progenitor cells or postmitotic neurons (Figure 3B and data not shown). However, although the number of TH+ dopaminergic neurons was unchanged (data not shown), notable differences between genotypes were seen. The number of neurons within the SNc that expressed the VTA marker Calbindin was markedly increased in E18.5 Sox6/ embryos (Figures 3C and 3F). The glycosylated form of DAT (GlycoDAT) was strongly expressed in SNc mDA neurons and in a dorsal-lateral population in the VTA in wildtype embryos. In contrast, neurons that were negative for GlycoDAT expression were significantly increased in Sox6/ SNc (Figures 3E and 3F), while expression of the Raldh1 in the SNc was reduced (Figures 3D and 3F).
In addition, at E18.5, striatal fiber innervation was affected in Sox6/ mutant embryos as shown by 3,30 -diaminobenzidine immunoperoxidase (DAB) staining for DAT and TH, respectively (Figures 3G and 3H; Figure S4). Notably, the lateral striatal region (area 1; Figure 3G), which is mainly innervated by SNc mDA neurons, was markedly less intensely stained as shown by densitometry using antibodies against both DAT and TH (Figure 3H; Figure S4). The more medial (area 2) and ventral striatum, which is predominantly innervated by VTA mDA neurons, showed normal TH and DAT immunostaining intensity (Figures 3G and 3H; Figure S4). In addition, expression of the ephrin receptor Epha5 was selectively expressed in SNc neurons in wild-type embryos and strongly reduced in Sox6/ animals, as determined by in situ hybridization analysis (Figure 3I). Thus, the reduced innervation of the lateral part of the striatum could be due to a defect in the axonal guidance pathway, but it may also be influenced by other deficiencies in SNc mDA neurons. The expression of several other axon-guidance molecules and
Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors 1021
general mDA neuron markers including Lmx1a, Pitx3, En1, and Nurr1 were unchanged (data not shown). Taken together, these data show that Otx2 plays a critical role in the suppression of Sox6 in mDA neuron progenitors and postmitotic neurons. Moreover, although a complete transition in phenotype from SNc to VTA neurons does not occur as a consequence of Sox6 ablation, the results clearly indicate that Sox6 plays an important role in promoting SNc-specific and suppressing VTA-specific characteristics in mDA neurons. Sox6 Is Essential in Postmitotic SNc Neurons and Is Expressed in Human SNc mDA Neurons Sox6/ mice do not survive after birth because of deficiencies that are unrelated to the phenotype in mDA neurons (Smits et al., 2001). To analyze the role of Sox6 in postmitotic mDA neurons, floxed Sox6 mice were crossed with DatCre mice to selectively inactivate Sox6 in Dat-expressing postmitotic mDA neurons (Kadkhodaei et al., 2009). Within the ventral midbrain, the number of mDA neurons appeared intact in DatCre/+;Sox6fl/fl mutant embryos and adult mice (data not shown). However, + there was a reduced density of dendritic TH fibers in the ventral midbrain of 4- and 8-month-old mutant mice (Figure 4D). In addition, while the striatal innervation in E18.5 DatCre/+;Sox6fl/fl mutant embryos appeared normal (Figure S3), TH+ fiber density in the striatum of 4-month-old mutant mice was clearly reduced (area 1; Figures 4A and 4C; Figure S4). The intensity of mDA neuron fiber density was further decreased in the dorsal-lateral part of the striatum of DatCre/+;Sox6fl/fl in 8-month-old animals (Figures 4B and 4C; Figure S4), while the innervation of the ventral striatum remained unaffected (Figures 4A–4C). Importantly, the affected striatal region is primarily innervated from Sox6+ SNc mDA neurons. Other analyzed regions within the hippocampus and amygdala appeared unaffected (data not shown). In addition, TH+ cell bodies within the VTA in 8-monthold mutant mice also appeared normal (Figure S4). Furthermore, consistent with a decreased striatal fiber innervation, DA and DA metabolites (DOPAC and HVA) were reduced in the caudate putamen (CPu) of Sox6 conditionally ablated mice (Figure 4E), while levels within VTA-innervated nucleus accumbens were not significantly altered (Figure S4). Further, the open field test showed that Sox6 conditional knockout mice were more active compared to controls (Figure S4), a behavior that has previously been shown to result from prenatal SNc mDA neuron deficiency (e.g., Hwang et al., 2005). Thus, ablation of Sox6 within differentiating mDA neurons demonstrates an important role in maintaining SNc, but not VTA, characteristics in mDA neurons. The results showing that Sox6 is important for maintaining normal postmitotic SNc mDA neuron properties in mice raise the question of how Sox6 is expressed in the normal and diseased human brain. Postmortem tissue samples from both Parkinson’s disease brains and aged-matched controls were analyzed by Sox6 immunostaining (Figures 4F–4H). Sox6 staining was clearly detectable and localized to neuromelanin and TH-positive neurons in the SNc (Figure 4F; Figure S4). Sox6 expression was not observed in medial neuromelanin-negative neurons (data not shown), but we could not firmly conclude whether these neurons are VTA neurons. Thus, Sox6 expression in the human mDA neurons is likely recapitulating that seen in
mice. To analyze whether Sox6 expression is affected in Parkinson’s disease, we compared the levels of Sox6 expression by densitometry in 145 individual neuromelanin-positive cells in controls (n = 4) and 323 neuromelanin-positive cells in Parkinson’s disease tissue samples (n = 8). Interestingly, a significant reduction of Sox6 staining in Parkinson’s disease patients compared to controls was evident when considering measurements in all individual neurons (Figure 4G) and averages in staining intensities determined for each control and each Parkinson’s disease patient (Figure 4H). Thus, Sox6 expression is clearly diminished in neuromelanin-positive neurons in Parkinson’s disease patients. DISCUSSION The importance of mDA neurons in normal brain function and their degeneration in Parkinson’s disease has placed this cell population at central stage in studies of lineage-specific neurogenesis during CNS development. An important motivation to understand how these cells develop originates from strong interest in developing methods for stem cell engineering of mDA neurons to be used in cell therapy or for disease modeling in cell culture (Cooper et al., 2012; Lindvall and Bjo¨rklund, 2011; Studer, 2012). However, although signaling and transcription factor pathways controlling generic aspects of mDA neuron specification and differentiation are beginning to be understood, the processes regulating mDA neuron diversity have remained mostly unclear. In this study, we elucidated a transcription factor code that subdivides the Lmx1a+ progenitor pool in lateral (Otx2highNolz1+Sox6) and medial (Otx2lowNolz1Sox6+) domains, respectively. After cell-cycle exit, these transcription factors are expressed in a corresponding way in VTA (Otx2+Nolz1+Sox6) and SNc (OtxNolz1Sox6+) postmitotic mDA neurons. This coordinated expression before and after cell-cycle exit suggests that subtype identities have already been determined in mDA neural progenitors. The finding that Sox6+ progenitor cells reside in medial domain while Nolz1+ and Otx2high progenitors are found in a lateral Sox6 region also indicated an intriguing, and somewhat counterintuitive, inside-out migration of developing SNc mDA neurons (Figure 2R). Interestingly, recent fate-mapping studies using temporally controlled Cre expression in reporter mice support the conclusion that SNc neurons originate from medial mDA progenitor cells (Blaess et al., 2011; Bodea et al., 2014; Joksimovic et al., 2009). Using two different mouse ablation strategies, we demonstrated that Sox6 is required for the normal development of SNc neurons and maintenance of some SNc-specific neuronal properties. Together, the data reveal both early and late roles for Sox6 in SNc neurons. Our finding that Sox6 ablation by DatCre leads to decreased density of TH+ fibers in the dorsal-lateral striatum and a corresponding decrease in the levels of striatal DA is intriguing when considering our observation that Sox6 expression is diminished in neuromelanin-positive neurons in Parkinson’s disease patients. Interestingly, a recent study identified a mutation within the human Sox6 gene leading to a general developmental delay and parkinsonian symptoms in an affected child (Scott et al., 2014). A decrease in Sox6 in association with
1022 Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors
B
4 months DatCre;Sox6
fl/+
DatCre;Sox6
1
fl/fl
*
**
1
2
2
2
area 1
area 2 area 1
TH-DAB
DatCre;Sox6fl/fl
DatCre;Sox6fl/+
DatCre;Sox6fl/+
area 2
DatCre;Sox6fl/fl
TH-DAB
D
Control DatCre;Sox6fl/fl
100%
1
TH-DAB
4 months 8 months
DatCre;Sox6fl/fl
DatCre;Sox6fl/+
1
2
C
8 months
Relative fiber density
A
**
p=0.052
*
Control DatCre;Sox6fl/fl
Sox6neuromelanin
F
Relative levels
100%
DA
H
DOPAC HVA
CPu Control 2
150 100 50 0
PD 1
PD 2
Control
PD
150
Densitometry (AU)
Control 1
***
200
G Densitometry (AU)
E
100
*
50
0
Control
PD
Figure 4. Sox6 Is Required for the Functional Integrity of SNc Neurons in Adult Mice (A) Immunohistochemical analysis of TH+ fiber innervation in the striatum of 4-month-old DatCre;Sox6fl/+ and DatCre;Sox6fl/fl mice. Area 1 represents the dorsallateral part of the striatum, while area 2 represents the ventral part of the striatum. (B) Immunohistochemical analysis of TH expression in the striatum of 8-month-old DatCre;Sox6fl/+ and DatCre;Sox6fl/fl mice. (C) Analysis of TH-expressing fibers in the striatum by densitometry in a dorsolateral (area1) and medial (area 2) region of the striatum at 4 and 8 months in DatCre;Sox6fl/+ and DatCre;Sox6fl/fl mice, respectively. (n = 4); mean values ± SD; *p < 0.05; **p < 0.01. (D) Analysis of TH expression in midbrain of 4- and 8-month-old DatCre;Sox6fl/+ and DatCre;Sox6fl/fl mice. Scale bars in (A)–(D), 200 mm. (E) DA and DA metabolites (DOPAC and HVA) analyzed by high-performance liquid chromatography from extracts of the dorsolateral part of the striatum (area 1) of 6-month-old DatCre;Sox6fl/+ and DatCre;Sox6fl/fl mice (n = 10). Values are shown as percentages relative the levels detected in extracts from DatCre;Sox6fl/+ mice (100%); mean values ± SD; *p < 0.05; **p < 0.01. (F) Sox6 expression in the human SNc. Expression analyzed by immunohistochemistry for Sox6 is shown in sections from two controls and two individuals with Parkinson’s disease (PD1 and PD2). Distribution of Sox6 (dark blue) is localized to neuromelanin-positive (brown) neurons in the SNc of control individuals (arrows), while nuclear Sox6 expression is not observed in PD patient 1 and only weakly observed in some nuclei of PD patient 2 (arrowheads). Scale bar, 25 mm. (G) Intensity of Sox6 levels in individual neuromelanin-positive neurons (n = 323) is decreased in in Parkinson’s disease patients compared to neurons (n = 145) in controls. Intensity was measured by densitometry and shown for each cell in arbitrary units; mean values ± SEM; ***p < 0.001. (H) The intensities from the individual cells (shown in G) is averaged for each Parkinson’s disease patient (n = 8) or control (n = 4) and determined in arbitrary units; mean values ± SEM; *p < 0.05. See also Figure S4.
Parkinson’s disease pathology may thus contribute to continuing deterioration of SNc mDA neurons in this disease. From the data reported here and from previous studies on Otx2, it should be noted that neither Otx2 nor Sox6 ablation led to a complete cell-fate transition of subtype identities (Di Giovan-
nantonio et al., 2013; Di Salvio et al., 2010b). Thus, we postulate that additional factors contribute to lineage selection of VTA and SNc neurons. It is interesting that Nolz1 is expressed strictly in lateral neural progenitors in which Otx2 is also highly expressed. An intriguing possibility is that Otx2 and Nolz1 functionally
Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors 1023
interact to determine the VTA neuronal fate. It is also important to note that the diversity of mDA neuron types extends beyond the division into SNc and VTA neurons. Thus, another major mDA neuron group localized in the retrorubral field (A8) has properties that are distinct from those in the SNc and VTA (Bjo¨rklund and Dunnett, 2007), and within the substantia nigra and VTA, molecularly and functionally diverse mDA neuron subpopulations have also been described (see e.g. Di Salvio et al., 2010a; Fu et al., 2011; Roeper, 2013). How Sox6, Otx2, Nolz1, and other transcription factors influence additional complexity within mDA neurons therefore remains an important question to investigate. The studies presented here have important implications for stem cell engineering. Cell transplantation in Parkinson’s disease animal models has demonstrated that Girk2+ SNc neurons must be a major constituent of cell grafts to restore lost motor function in Parkinson’s disease (Cooper et al., 2012; Grealish et al., 2010; Zuddas et al., 1991). The identification of Sox6 and Nolz1 as two early progenitor markers for prospective SNc and VTA neurons, respectively, will provide a valuable tool in tracing how different culturing conditions affect lineage selection at very early stages in stem cell differentiation experiments. This, in turn, should facilitate possibilities to identify culturing conditions resulting in SNc or VTA mDA neuron-enriched cultures (Studer, 2012). EXPERIMENTAL PROCEDURES Mouse Strains Sox6fl/+ animals (Dumitriu et al., 2006) were crossed with DatCre/+;Sox6fl/+ animals (Ekstrand et al., 2007) to obtain DatCre/+;Sox6fl/fl mutant offspring. Sox6/ mutant embryos were obtained from crossings between Sox6+/ heterozygous mice. Otx2 mutant mouse strains were kept and bred as described previously (Di Giovannantonio et al., 2013; Di Salvio et al., 2010a). All animal experiments were approved by the local animal ethics committee and conform to the relevant regulatory standards. For details, see Supplemental Experimental Procedures. Immunohistochemistry Embryos were fixed for 1–3 hr in 4% paraformaldehyde (PFA) and further processed as described before (Briscoe et al., 2000). Adult mice were perfused with 4% PFA as described previously (Kadkhodaei et al., 2009), and dissected brains were postfixed for 4 hr in 4%PFA. For details, see Supplemental Experimental Procedures. Human Postmortem Tissue Paraffin-embedded midbrain sections from Parkinson’s disease patients (n = 8) and age-matched control individuals (n = 4) were provided by the UK Parkinson’s Disease Society Tissue Bank at Imperial College. For details, see Supplemental Experimental Procedures. Statistical Analysis Statistical significance was calculated by the one-paired Student’s t test, and data are presented as mean ± SD (*p < 0.05, **p < 0.01). SUPPLEMENTAL INFORMATION
experiment; A.L. analyzed human tissue; T.O. performed histology; D.A. and A.S. analyzed Otx2 ablated embryos; N.V. helped write the manuscript; I.K. bred mice; T.Y. and J.K. performed high-performance liquid chromatography analysis; E.J. performed histology; J.M. and J.E. helped with analysis and planning; and T.P. together with L.P. planned all experiments and wrote the manuscript. ACKNOWLEDGMENTS We thank Bhumica Singla and Hilda Lunden-Miguel for technical assistance. We thank Veronique Lefebvre and Meng Li for generously providing mouse lines. This work was supported by funding from the European Union, Seventh Framework Programme under grant agreement mdDANeurodev and NeuroStemCell (to T.P.), from the Swedish Strategic Research Foundation (SSF; to T.P.), and from Vetenskapsra˚det (L.P. and T.P.) and Parkinsonfonden (L.P.) Received: December 27, 2013 Revised: June 24, 2014 Accepted: July 14, 2014 Published: August 7, 2014 REFERENCES Andersson, E., Tryggvason, U., Deng, Q., Friling, S., Alekseenko, Z., Robert, B., Perlmann, T., and Ericson, J. (2006). Identification of intrinsic determinants of midbrain dopamine neurons. Cell 124, 393–405. Azim, E., Jabaudon, D., Fame, R.M., and Macklis, J.D. (2009). SOX6 controls dorsal progenitor identity and interneuron diversity during neocortical development. Nat. Neurosci. 12, 1238–1247. Batista-Brito, R., Rossignol, E., Hjerling-Leffler, J., Denaxa, M., Wegner, M., Lefebvre, V., Pachnis, V., and Fishell, G. (2009). The cell-intrinsic requirement of Sox6 for cortical interneuron development. Neuron 63, 466–481. Bjo¨rklund, A., and Dunnett, S.B. (2007). Dopamine neuron systems in the brain: an update. Trends Neurosci. 30, 194–202. Blaess, S., Bodea, G.O., Kabanova, A., Chanet, S., Mugniery, E., Derouiche, A., Stephen, D., and Joyner, A.L. (2011). Temporal-spatial changes in Sonic Hedgehog expression and signaling reveal different potentials of ventral mesencephalic progenitors to populate distinct ventral midbrain nuclei. Neural Dev. 6, 29. Bodea, G.O., Spille, J.H., Abe, P., Andersson, A.S., Acker-Palmer, A., Stumm, R., Kubitscheck, U., and Blaess, S. (2014). Reelin and CXCL12 regulate distinct migratory behaviors during the development of the dopaminergic system. Development 141, 661–673. Briscoe, J., Pierani, A., Jessell, T.M., and Ericson, J. (2000). A homeodomain protein code specifies progenitor cell identity and neuronal fate in the ventral neural tube. Cell 101, 435–445. Castelo-Branco, G., Wagner, J., Rodriguez, F.J., Kele, J., Sousa, K., Rawal, N., Pasolli, H.A., Fuchs, E., Kitajewski, J., and Arenas, E. (2003). Differential regulation of midbrain dopaminergic neuron development by Wnt-1, Wnt-3a, and Wnt-5a. Proc. Natl. Acad. Sci. USA 100, 12747–12752. Chang, S.L.-Y., Chen, S.-Y., Huang, H.-H., Ko, H.-A., Liu, P.-T., Liu, Y.-C., Chen, P.-H., and Liu, F.-C. (2013). Ectopic expression of nolz-1 in neural progenitors promotes cell cycle exit/premature neuronal differentiation accompanying with abnormal apoptosis in the developing mouse telencephalon. PLoS ONE 8, e74975.
Supplemental Information includes Supplemental Experimental Procedures and four figures and can be found with this article online at http://dx.doi.org/ 10.1016/j.celrep.2014.07.016.
Chung, C.Y., Seo, H., Sonntag, K.C., Brooks, A., Lin, L., and Isacson, O. (2005). Cell type-specific gene expression of midbrain dopaminergic neurons reveals molecules involved in their vulnerability and protection. Hum. Mol. Genet. 14, 1709–1725.
AUTHOR CONTRIBUTIONS
Chung, C.Y., Licznerski, P., Alavian, K.N., Simeone, A., Lin, Z., Martin, E., Vance, J., and Isacson, O. (2010). The transcription factor orthodenticle homeobox 2 influences axonal projections and vulnerability of midbrain dopaminergic neurons. Brain 133, 2022–2031.
L.P. planned all experiments, performed histological analysis, and wrote the manuscript; M.P. performed behavioral analysis and the fluorogold tracing
1024 Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors
Cooper, O., Parmar, M., and Isacson, O. (2012). Characterization and criteria of embryonic stem and induced pluripotent stem cells for a dopamine replacement therapy. Prog. Brain Res. 200, 265–276. Damier, P., Hirsch, E.C., Agid, Y., and Graybiel, A.M. (1999). The substantia nigra of the human brain. II. Patterns of loss of dopamine-containing neurons in Parkinson’s disease. Brain 122, 1437–1448. Di Giovannantonio, L.G., Di Salvio, M., Acampora, D., Prakash, N., Wurst, W., and Simeone, A. (2013). Otx2 selectively controls the neurogenesis of specific neuronal subtypes of the ventral tegmental area and compensates En1dependent neuronal loss and MPTP vulnerability. Dev. Biol. 373, 176–183.
Nurr1 is required for maintenance of maturing and adult midbrain dopamine neurons. J. Neurosci. 29, 15923–15932. Lindvall, O., and Bjo¨rklund, A. (2011). Cell therapeutics in Parkinson’s disease. Neurotherapeutics 8, 539–548. Marı´n, F., Herrero, M.-T., Vyas, S., and Puelles, L. (2005). Ontogeny of tyrosine hydroxylase mRNA expression in mid- and forebrain: neuromeric pattern and novel positive regions. Dev. Dyn. 234, 709–717. McRitchie, D.A., Hardman, C.D., and Halliday, G.M. (1996). Cytoarchitectural distribution of calcium binding proteins in midbrain dopaminergic regions of rats and humans. J. Comp. Neurol. 364, 121–150.
Di Salvio, M., Di Giovannantonio, L.G., Omodei, D., Acampora, D., and Simeone, A. (2010a). Otx2 expression is restricted to dopaminergic neurons of the ventral tegmental area in the adult brain. Int. J. Dev. Biol. 54, 939–945.
Nunes, I., Tovmasian, L.T., Silva, R.M., Burke, R.E., and Goff, S.P. (2003). Pitx3 is required for development of substantia nigra dopaminergic neurons. Proc. Natl. Acad. Sci. USA 100, 4245–4250.
Di Salvio, M., Di Giovannantonio, L.G., Acampora, D., Prosperi, R., Omodei, D., Prakash, N., Wurst, W., and Simeone, A. (2010b). Otx2 controls neuron subtype identity in ventral tegmental area and antagonizes vulnerability to MPTP. Nat. Neurosci. 13, 1481–1488.
Panman, L., Andersson, E., Alekseenko, Z., Hedlund, E., Kee, N., Mong, J., Uhde, C.W., Deng, Q., Sandberg, R., Stanton, L.W., et al. (2011). Transcription factor-induced lineage selection of stem-cell-derived neural progenitor cells. Cell Stem Cell 8, 663–675.
Dumitriu, B., Dy, P., Smits, P., and Lefebvre, V. (2006). Generation of mice harboring a Sox6 conditional null allele. Genesis 44, 219–224.
Prakash, N., Brodski, C., Naserke, T., Puelles, E., Gogoi, R., Hall, A., Panhuysen, M., Echevarria, D., Sussel, L., Weisenhorn, D.M.V., et al. (2006). A Wnt1-regulated genetic network controls the identity and fate of midbraindopaminergic progenitors in vivo. Development 133, 89–98.
Ekstrand, M.I., Terzioglu, M., Galter, D., Zhu, S., Hofstetter, C., Lindqvist, E., Thams, S., Bergstrand, A., Hansson, F.S., Trifunovic, A., et al. (2007). Progressive parkinsonism in mice with respiratory-chain-deficient dopamine neurons. Proc. Natl. Acad. Sci. USA 104, 1325–1330. Fearnley, J.M., and Lees, A.J. (1991). Ageing and Parkinson’s disease: substantia nigra regional selectivity. Brain 114, 2283–2301. Ferri, A.L.M., Lin, W., Mavromatakis, Y.E., Wang, J.C., Sasaki, H., Whitsett, J.A., and Ang, S.-L. (2007). Foxa1 and Foxa2 regulate multiple phases of midbrain dopaminergic neuron development in a dosage-dependent manner. Development 134, 2761–2769. Fu, Y., Yuan, Y., Halliday, G., Ruszna´k, Z., Watson, C., and Paxinos, G. (2011). A cytoarchitectonic and chemoarchitectonic analysis of the dopamine cell groups in the substantia nigra, ventral tegmental area, and retrorubral field in the mouse. Brain Struct. Funct. 217, 591–612. Grealish, S., Jo¨nsson, M.E., Li, M., Kirik, D., Bjo¨rklund, A., and Thompson, L.H. (2010). The A9 dopamine neuron component in grafts of ventral mesencephalon is an important determinant for recovery of motor function in a rat model of Parkinson’s disease. Brain 133, 482–495. Hwang, D.-Y., Ardayfio, P., Kang, U.J., Semina, E.V., and Kim, K.-S. (2003). Selective loss of dopaminergic neurons in the substantia nigra of Pitx3-deficient aphakia mice. Brain Res. Mol. Brain Res. 114, 123–131. Hwang, D.-Y., Fleming, S.M., Ardayfio, P., Moran-Gates, T., Kim, H., Tarazi, F.I., Chesselet, M.-F., and Kim, K.-S. (2005). 3,4-dihydroxyphenylalanine reverses the motor deficits in Pitx3-deficient aphakia mice: behavioral characterization of a novel genetic model of Parkinson’s disease. J. Neurosci. 25, 2132–2137. Ji, S.-J., Periz, G., and Sockanathan, S. (2009). Nolz1 is induced by retinoid signals and controls motoneuron subtype identity through distinct repressor activities. Development 136, 231–240. Joksimovic, M., Anderegg, A., Roy, A., Campochiaro, L., Yun, B., Kittappa, R., McKay, R., and Awatramani, R. (2009). Spatiotemporally separable Shh domains in the midbrain define distinct dopaminergic progenitor pools. Proc. Natl. Acad. Sci. USA 106, 19185–19190. Kadkhodaei, B., Ito, T., Joodmardi, E., Mattsson, B., Rouillard, C., Carta, M., Muramatsu, S., Sumi-Ichinose, C., Nomura, T., Metzger, D., et al. (2009).
Puelles, E., Annino, A., Tuorto, F., Usiello, A., Acampora, D., Czerny, T., Brodski, C., Ang, S.-L., Wurst, W., and Simeone, A. (2004). Otx2 regulates the extent, identity and fate of neuronal progenitor domains in the ventral midbrain. Development 131, 2037–2048. Roeper, J. (2013). Dissecting the diversity of midbrain dopamine neurons. Trends Neurosci. 36, 336–342. Scott, O., Pugh, J., Kiddoo, D., Sonnenberg, L.K., Bamforth, S., and Goez, H.R. (2014). Global developmental delay, progressive relapsing-remitting parkinsonism, and spinal syrinx in a child with SOX6 mutation. J. Child Neurol., Published online January 21, 2014 http://dx.doi.org/10.1177/0883073813514134. Smidt, M.P., Asbreuk, C.H., Cox, J.J., Chen, H., Johnson, R.L., and Burbach, J.P. (2000). A second independent pathway for development of mesencephalic dopaminergic neurons requires Lmx1b. Nat. Neurosci. 3, 337–341. Smidt, M.P., Smits, S.M., Bouwmeester, H., Hamers, F.P.T., van der Linden, A.J.A., Hellemons, A.J.C.G.M., Graw, J., and Burbach, J.P.H. (2004). Early developmental failure of substantia nigra dopamine neurons in mice lacking the homeodomain gene Pitx3. Development 131, 1145–1155. Smits, P., Li, P., Mandel, J., Zhang, Z., Deng, J.M., Behringer, R.R., de Crombrugghe, B., and Lefebvre, V. (2001). The transcription factors L-Sox5 and Sox6 are essential for cartilage formation. Dev. Cell 1, 277–290. Studer, L. (2012). Derivation of dopaminergic neurons from pluripotent stem cells. Prog. Brain Res. 200, 243–263. van den Munckhof, P., Luk, K.C., Ste-Marie, L., Montgomery, J., Blanchet, P.J., Sadikot, A.F., and Drouin, J. (2003). Pitx3 is required for motor activity and for survival of a subset of midbrain dopaminergic neurons. Development 130, 2535–2542. Wolfart, J., Neuhoff, H., Franz, O., and Roeper, J. (2001). Differential expression of the small-conductance, calcium-activated potassium channel SK3 is critical for pacemaker control in dopaminergic midbrain neurons. J. Neurosci. 21, 3443–3456. Zuddas, A., Corsini, G.U., Barker, J.L., Kopin, I.J., and Di Porzio, U. (1991). Specific Reinnervation of Lesioned Mouse Striatum by Grafted Mesencephalic Dopaminergic Neurons. Eur. J. Neurosci. 3, 72–85.
Cell Reports 8, 1018–1025, August 21, 2014 ª2014 The Authors 1025
