https://orcid.org/0000-0003-3075-1553
Figures
Abstract
In this study we investigate how β-catenin-dependent WNT signalling impacts midbrain dopaminergic neuron (mDA) specification. mDA cultures at day 65 of differentiation responded to 25 days of the tankyrase inhibitor XAV969 (XAV, 100nM) with reduced expression of markers of an A9 mDA phenotype (KCNJ6, ALDH1A1 and TH) but increased expression of the transcriptional repressors NR0B1 and NR0B2. Overexpression of NR0B1 and or NR0B2 promoted a loss of A9 dopaminergic neuron phenotype markers (KCNJ6, ALDH1A1 and TH). Overexpression of NR0B1, but not NR0B2 promoted a reduction in expression of the β-catenin-dependent WNT signalling pathway activator RSPO2. Analysis of Parkinson’s disease (PD) transcriptomic databases shows a profound PD-associated elevation of NR0B1 as well as reduced transcript for RSPO2. We conclude that reduced β-catenin-dependent WNT signalling impacts dopaminergic neuron identity, in vitro, through increased expression of the transcriptional repressor, NR0B1. We also speculate that dopaminergic neuron regulatory mechanisms may be perturbed in PD and that this may have an impact upon both existing nigral neurons and also neural progenitors transplanted as PD therapy.
Citation: Haynes JM, Sibuea SM, Aguiar AA, Li F, Ho JK, Pouton CW (2021) Inhibition of β-catenin dependent WNT signalling upregulates the transcriptional repressor NR0B1 and downregulates markers of an A9 phenotype in human embryonic stem cell-derived dopaminergic neurons: Implications for Parkinson’s disease. PLoS ONE 16(12): e0261730. https://doi.org/10.1371/journal.pone.0261730
Editor: Renping Zhou, Rutgers University, UNITED STATES
Received: September 16, 2021; Accepted: December 7, 2021; Published: December 23, 2021
Copyright: © 2021 Haynes et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files. This includes a list to all relevant GEO files utilized.
Funding: Parts of this work were funded by the Yulgilbar Alzheimer’s Research Program, Stem Cells Australia (SR1101002) and the Australian Research Council (DP140103743). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Parkinson’s disease is a debilitating disease affecting 1–2% of individuals over 50 years of age [1]. Patients with Parkinson’s disease suffer a variety of well-documented motor and non-motor symptoms. The motor symptoms include hypokinesia, muscle rigidity and shaking tremor [2], while the non-motor symptoms include depression, apathy, anxiety, hallucinations, sleep disorders, urinary urgency, nocturia, sexual dysfunction, dysphagia, fecal incontinence and paresthesia [3]. The classically described motor symptoms of Parkinson’s disease result from the loss of tyrosine hydroxylase (TH) positive dopaminergic (A9) neurons of the substantia nigra pars compacta. This results in a deficit of striatal dopamine and impaired motor function. Current pharmacotherapies encourage production of, or mimic or modulate the post-synaptic effects of dopamine. Newer therapies target increased survival of remaining neurons [4], while large scale trials of dopaminergic neural progenitor therapy are planned or underway (E.g. European Stem-PD, NYSTEM-PD, CiRA trial).
We know a great deal about the derivation of dopaminergic neurons from our understanding of mouse development. As the embryo develops, midbrain dopaminergic neuron development is largely under the control of sonic hedgehog (Shh) and β-catenin-dependent Wnt signalling [5–7]. After patterning, β-catenin-independent Wnt signalling suppresses β-catenin-dependent Wnt signalling to facilitate differentiation and maturation events [8]. Human pluripotent stem cell differentiation protocols attempt to mimic this sequence of events such that, maturing mouse and also human dopaminergic neural progenitors show early expression of transcription factors such as EN1/2, OTX2, FOXA2, LMX1A, MSX1, NEUROG2 and late stage factors such as ASCL1, NR4A2, PITX3 [9–13]. At maturation, PITX3 expression is important in defining mature midbrain dopaminergic neurons since it is highly localized to the midbrain [14], particularly the ventral substantia nigra [15]. Although PITX3 is highly localized to the midbrain, the rate limiting enzyme responsible for the synthesis of dopamine, tyrosine hydroxylase (TH), is commonly used as the definitive marker of dopaminergic phenotype. At maturity, substantia nigra (A9) dopaminergic neurons can be identified by expression of KCNJ6 (G Protein-Activated Inward Rectifier Potassium Channel 2 [16]), aldehyde dehydrogenase (ALDH1A1, [17]) and SOX6 (SRY, Sex Determining Region Y-Box 6, [18]). Ventral tegmental area (A10) dopaminergic neurons are often identified with calcium binding protein (calbindin, CALB1) and / or OTX2 [16, 18]. Significantly, the WNT signalling inhibitor dickkopf-3 has been linked to both the maintenance of Lmx1a and Pitx3 expression in mice, and also to elements of subtype (A9 v A10) specification [19]. The role of WNTs in regulating biological process is also expanding beyond developmental programming, thus WNT signalling may have a role to play in some genetic forms of PD [20] which is consistent with the idea that WNT signalling is involved in neuron survival [21, 22].
Transplantation therapies for PD favour differentiating dopaminergic neuronal progenitors (~ day 20) as they offer improved connectivity and integration with host neurons as they mature (for reviews see [23, 24]). To that end there has been a continual and iterative push to create better dopaminergic neuron differentiation paradigms and/or purify dopaminergic neuronal progenitors (for example, see [25, 26]). In contrast with our understanding of early dopaminergic progenitor development, we have little understanding of what supports dopaminergic phenotype later in development, or indeed, after transplantation into humans. In this study, we cultivate pluripotent stem cell derived neurons for 65 days in vitro, and assess their transcriptomic profiles following a 25-day (day 40–65) incubation with either the canonical WNT signalling mimetic CHIR99025 (CHIR) or the canonical WNT signalling inhibitor, XAV969 (XAV). We subsequently overexpress three genes of interest (NEUROD1, NR0B1 and NR0B2) and assess their impact upon dopaminergic phenotype. Our data indicate a complex WNT regulatory environment in maturing cultures and a profound regulation of dopaminergic neuronal phenotype, particularly by XAV. Analysis of PD transcriptomic databases indicates changes in endogenous WNT signalling accompanying the loss of neurons as well as markers of a dopaminergic A9 neuronal phenotype. These data lead us to suggest that, even after patterning, and at a time point beyond that at which they would be transplanted, maturing dopaminergic neurons are still strongly influenced by WNT signalling. This has major implications for the transplantation of neural progenitors for PD where the maintenance of dopaminergic phenotype during integration and beyond maturation may still be subject to influence by endogenous brain chemistry.
Materials and methods
Parkinson’s disease array databases: For this analysis we accessed 11 array database sets to compare substantia nigra from PD and age-matched control tissues. GEO accession numbers: GSE20163 [27], GSE20141 [27]; GSE20164 [28]; GSE20333 (Edna et al. [Unpublished]); GSE7307 [29]; GSE7621 [28]; GDS3128 [30]; GDS3129 [30]; GSE54282 [31]; GSE43490 [32]; GSE8397 [33]. These datasets used any of five Affymetrix (GPL201, GL570, GPL17047, GPL97 and GPL96) or Agilent (GPL6480) platforms. S1 Table includes the specific data sets used for this study. For each gene of interest, transcript expression was standardized to 12 housekeeper genes (TRFC, RPLP0, PGK1, PPIA, ACTN1, B2M, GUSB, HPRT1, STAT1, ACTB, GAPDH). Given the wide range of platforms used, we kept analysis to a simple paired t-test design where each set of PD and control values from a single database contributed to an n of 1. Where lateral vs medial nigral tissue was used (for example [33]), we have used the average of both lateral and nigral expression to maintain consistency with whole nigral studies.
Human ESC culture
The cell line used in this study expresses eGFP under the control of LMX1A (as described previously [34]). Briefly, the reporter hESC line (H9LMX1A/eGFP) was cultured, feeder free, on recombinant human laminin 521 (1μg cm-2, Life Technologies, Australia) using Essential 8 medium (Life Technologies, Australia). For hESC differentiation the method of [35] was employed. Briefly, on Day 0 of differentiation, hESCs were seeded onto Matrigel coated plates at a density of 120,000 cells cm-2, in 100% Midbrain Differentiation Media 1 (MDM1: Knock-out Serum Replacement containing: 15%, 1% Glutamax, 1% Non-Essential Amino Acids, 1% Penicillin/Streptomycin, 0.1% β-Mercaptoethanol, 100nM LDN-193189 and 10μM SB431542). On Day 1, media was replaced with MDM2: MDM1 containing sonic hedgehog (100ng/mL) and purmorphamine (2μM). On Day 3, media was replaced with MDM3: MDM2 containing CHIR99026 (CHIR, 3μM). After eight days (Day 11) the media was gradually replaced with maturation media (Neurobasal Medium containing 1% Penicillin/Streptomycin, 1% Glutamax, 2% B27 Supplement without Retinoic Acid as well as 20ng/mL each of BDNF and GDNF, 200μM Ascorbic Acid, 2.5μM DAPT, 1ng/mL TGFβ-3, 0.5mM dibutyryl cyclic AMP). The maturation media was completely replaced every two days. From days 40 to 65 of differentiation CHIR (1μM) or XAV939 (XAV, 100nM), or vehicle, were added to cultures. On the day of cells were harvested for quantitative reverse transcriptase PCR (qPCR) or immunocytochemistry.
Overexpression of NEUROD1 / NR0B1 / NR0B2
Cultures were differentiated as described above and matured until day 55. At this time, cultures were transfected with NEUROD1 (SC118625, Origene), NR0B1 (SC319624, Origene), NR0B2 (SC122920, Origene) or blank plasmids (PCMV6XL5, Origene) according to manufacturer’s instructions. At days 1, 2, 6, 7, 9,10 post-transfection, NEUROD1 transcript and protein levels were established using qPCR and immunolabelling (we used days 3, 6 and 9 for NR0B1/2).
qPCR
Quantitative PCR was undertaken using a method previously described [36]. RNA was extracted from 106 cells using ISOLATE II Micro Kit (Bioline) according to manufacturer’s instruction. Briefly, cell lysis buffer is added and cell lysate was briefly centrifuged at 1000xg for five seconds, followed by homogenization with 70% (v/v) ethanol. Homogenized lysate was transferred to a small column with silica membrane and centrifuged to allow RNA binding to the membrane. After the wash steps, the RNA was eluted with 15 μL RNase-free water. Total RNA concentration was measured via the NanoDrop Spectrophotometer (ND-1000, ThermoFisher Scientific). Agarose gel electrophoresis was used to confirm the absence of DNA. To convert RNA to complementary DNA (cDNA), a SensiFast cDNA Synthesis Kit (Bioline) was used. Briefly, the extracted RNA was transferred to a Polymerase Chain Reaction (PCR) tube with 1 μL Reverse Transcriptase and 4 μL TransAmp Buffer which was then brought up to 20 μL with RNase-free water. Tubes were placed in a thermal cycler (Applied Biosystem) set to the protocol stipulated in the Bioline manual. Converted cDNA samples were stored at -20°C until further analysis. Three technical replicate reactions were performed on samples aggregated from at least three independently differentiating wells using the Bioline Sensifast SYBR No-ROX One Step Kit according to the manufacturer’s specifications. Relative quantification of gene expression was obtained using the ratio Ct values of target genes to mean Ct values to housekeepers, TBP1 & HPRT1.The list of Taqman probes is shown in S2 Table while S3 Table shows delta Ct values for each replicate.
Immunocytochemistry and immunohistochemistry
Human brain sections from PD (three females aged 74, 65 and 64) and control subjects (three females aged 76, 67 and 64) were fixed to gelatin subbed microscope slides and fixed with 4% (w/v) paraformaldehyde (Merck, Australia) in PBS. Day 65 cultured neurons were fixed with 4% (w/v) paraformaldehyde (Merck, Australia) in PBS. Preparation for imaging was identical for both human tissues cultured cells. Samples were permeabilized with 0.05% (v/v) Triton™ X-100 (PBST; Sigma-Aldrich) in PBS for 20 minutes at room temperature and then blocked with 3% (v/v) normal donkey serum (Merck, Australia) in PBS for 20 minutes then incubated overnight in PBST with primary antibodies at 4°C (Abcam Australia, S4 Table). Samples were then incubated with secondary antibodies (Abcam Australia, S4 Table) diluted in PBST for 2 hours at room temperature prior to counterstaining with Hoechst 33342 (1:5000; Thermo Fisher Scientific). Fluorescence images were captured using a Nikon Ti A1R inverted confocal microscope (Nikon Instruments, Japan). Donor tissues were received from the Victorian Brain Bank: Project number 17.22. Histopathological examination of tissues confirmed diagnosis of PD in PD, but not control specimens. Monash University Human research ethics committee granted exemption from requiring ethics approval since deidentified human tissue sections were obtained from the Victorian brain bank network.
Putative transcription factor binding sites
To better understand the differential regulation of genes by WNT signalling modulators and NEUROD1 we used MotifMap [37], to identify potential gene regulatory motifs an arbitrary 10,000 base pairs up and 10,000 base pairs downstream of transcription start sites. Later, we used two other web-based tools to assess agreement between different platforms in defining putative transcription factor binding sites: ConTra V3 [38] and LASAGNA [39]. Broadly speaking, the results across the three platforms were largely compatible although the LASAGNA and MotifMap showed more consistency in identify and quantifying numbers of putative transcription factor binding sites.
Statistical analyses
For most qPCR we used repeated measures one-way ANOVA with post-hoc Dunnett’s test to identify differences in Δct values. Where qPCR showed undetectable expression (ie. >40 cycles) the non-parametric Kruskal-Wallis test with post-hoc Dunn’s test was used (where undetectable values were ranked as equal lowest expression). For expression in graphs, data sets are shown as log2 Δct or log2 ΔΔct. For NEUROD1 and NR0B1/2 overexpression studies Student’s paired t-tests, or repeated measures two-way ANOVA with post-hoc Dunnett’s test were used where times after transfection (not including t = 0) were arranged in columns and analysis identified main column effect (ie. differences in expression over the time course of the study). All statistical analyses were performed with GraphPad Prism 8.
Results
Expression profiles of developing cultures change over time
We used qPCR to track the fate of dopaminergic neuronal cultures from day 20 to day 80 of differentiation. There were significant decreases in LIM Homeobox Transcription Factor 1α (LMX1A) and Forkhead Box A2 (FOXA2) over time, although LMX1A expression appeared to plateau from day 40. There were also elevations of α-synuclein (SCNA), Solute Carrier Family 18 Member A2 (SLC18A2 or VMAT2) and tyrosine hydroxylase (TH) (Fig 1). These results are consistent with the idea that maturing cultures reduce expression of early markers, such as FOXA2 and LMX1A while increasing expression of markers associated with a more mature dopaminergic neuronal phenotype, TH, SLC18A2 and SCNA. The expression of Nuclear Receptor Subfamily 4 Group A Member 2 (NR4A2), a transcription factor associated with dopaminergic neurogenesis and maintenance, peaked at day 40 before slowly reducing over time. Developing cultures also showed a relatively constant expression of β3-tubulin (TUBB3), but an elevation of glial associated fibrillary acid (GFAP) transcript over time, indicating that astrocyte like cells develop after neurons.
Comparison of dopaminergic gene markers gene expression at days 20,40, 60 and 80 of differentiation. Data are expressed as log2 ΔCt values for presentation. Most statistical analysis used One-way ANOVA followed by Dunnett’s multiple comparison test, day 20 was used for comparison, (n = 3–26). *, ** and *** indicate P<0.05, <0.01 and <0.001, respectively. Since GFAP was not detected at day 20 (Un, >40 cycles) and not measured at day 80, we used a Student’s t-test to compare values at day 40 and 60.
We then assessed the impact of the β-catenin-dependent WNT signalling modulators CHIR and XAV upon cultures, late in differentiation during a period of relatively stable gene expression (days 40–65). CHIR (1μM) reduced expression of two gene transcripts associated with dopaminergic neuron phenotype; NR4A2 and TH without greatly impacting others (Fig 2). In contrast, XAV elicited significant reductions in multiple markers of a dopaminergic (and in particular A9) phenotype: SLC18A2, Potassium Inwardly Rectifying Channel Subfamily J Member 6 (KCNJ6 or GIRK2), Aldehyde Dehydrogenase 1 Family Member A1 (ALDH1A1), TH, Paired Like Homeodomain 3 (PITX3), LMX1A, Orthodenticle Homeobox 2 (OTX2), and EN2. XAV also increased expression of another A9 marker, SRY-Box Transcription Factor 6 (SOX6), but did not change expression of the generic neuronal markers β3 tubulin or synuclein. XAV also decreased expression of nestin, and WNT5A, but not WNT1 (Fig 2). Immunolabelling of cultures with antibodies for TH, KCNJ6 and SLC6A3 was consistent with PCR studies, showing some changes following incubations with both CHIR (TH and KCNJ6) and XAV (TH, SLC6A3 and KCNJ6; Fig 3, panels a, b and c, respectively).
Data are expressed as log2 ΔΔCt values for presentation, but statistical analysis was performed using XAV or CHIR vs vehicle control ΔCt values in either paired t-tests or repeated measures one-ANOVA with post-hoc Dunnett’s tests (n = 3–10). *and ** indicate P<0.05 and <0.01, respectively.
Panel (a) shows tyrosine hydroxylase (TH), panel (b) shows dopamine transporter (SLC6A3) and panel (c) shows potassium inwardly rectifying channel subfamily J member 6 (KCNJ6) in maturing cultures directed toward midbrain differentiation (day 65). Within panels (a), (b) and (c) each of the images, from left to right, show DAPI, LMX1A (eGFP), gene of interest (TH, SLC6A3 or KCNJ6) and a colour combined image. Scale bar indicates 100μm.
Another transcription factor affected by XAV was Neuronal Differentiation 1 (NEUROD1), an important regulator of neuronal differentiation expressed within the midbrain area throughout life, and overlapping with TH expression [40].
Is there a common transcription factor regulating genes affected by XAV?
We turned to MotifMap [37] to look for common putative transcription factor binding sites in the genes whose expression was reduced by XAV (ie. ALDH1A1, SLC18A2, TH, KCNJ6, EN2, PITX3 LMX1A, OTX2, NES and WNT5A, Fig 2). This analysis revealed a short list of putative common transcription factor binding sites (in order of abundance, but not corrected for orientation): NEURO D, LEF1, PDX1 (IPF1), TEAD1, MAFB, ETS2, NR4A2 and STATs (including STAT1,2,3,4,5A & 6). A complete list of putative transcription factor binding sites as indicated by MotifMap is shown in S5 Table.
While the preponderance of putative LEF1 transcription factor binding sites in genes affected by XAV was reassuring, some genes showing profound effects, namely EN2 and PITX3 possessed few putative LEF1 transcription factor binding sites but were rich with putative NEURO D transcription factor binding sites. That XAV (100nM) reduced NEUROD1 expression in cultures (Fig 2) is therefore consistent with the idea that a loss of NEUROD1 signalling contributes to a loss of dopaminergic neuronal phenotype. Given the early and persistent expression of NEUROD1 in mouse dopaminergic neuron differentiation and maturation [40], we began with the idea that NEUROD1 overexpression might serve to rescue dopaminergic phenotype under the influence of XAV signalling. Surprisingly, our initial experiments showed that following transfection, cultures overexpressing NEUROD1 showed reduced ALDH1A1, TH, KCNJ6 and LMX1A, along with no change in NR4A2, SLC6A3 SLC18A2 or TUBB3 transcript (Fig 4; immunolabelling showing increased NEUROD1 protein is shown in S1 Fig). Essentially, this data indicates that while neuronal phenotype was largely unaffected, NEUROD1 reduced expression of markers of an A9 dopaminergic neuron phenotype (i.e. KCNJ6, ALDH1A1 and TH). These data are more in agreement with the findings of Park et al., [41] that NEUROD1 opposes the impact of NR4A2, rather than Val Cervo et al., [42] who showed that NEUROD1 promoted a dopaminergic neuronal phenotype. Although, our data indicate that overexpression of NEUROD1 reduces dopaminergic, and in particular, an A9 phenotype we also saw that NEUROD1 overexpression did not affect WNT1 or WNT5A expression (not shown). Given that NEUROD1 is widely regarded as a transcriptional activator, we searched the literature for potential mechanisms that might enable NEUROD1 to inhibit expression of ALDH1A1, KCNJ6, TH and LMX1A. We identified SHP1 (small heterodimer partner 1, NR0B2), a corepressor of NEURO D (NEUROD1) that competes with coactivator p300 for binding sites [43], and another repressor protein DAX (NR0B1) [44] that represses the activity of another gene; hepatocyte nuclear factor 4 (HNF4) [45, 46], previously identified as a potential biomarker for PD [47].
Cultures showed increased NEUROD1 along with reduced ALDH1A1, TH, KCNJ6 and LMX1A, but no change in NR4A2, SLC6A3 SLC18A2 or TUBB3. Control wells underwent transfection with an empty plasmid. Results are expressed as mean ± SEM of three (biological) replicate experiments. Analysis was performed with t-tests where each post transfection timepoint represented a paired measure (t = 0 was not included in analyses). * and ** indicate P<0.05 and P<0.01, respectively. S1 Fig shows NEUROD1 immunolabelling following transfection.
We assessed the activity of XAV and CHIR in regulating NR0B1 & NR0B2 expression and found that XAV, but not CHIR, significantly elevated both transcripts (Fig 5, panels (a) and (c)). Consistent with the impacts upon transcript expression, NR0B1 and NR0B2 proteins were largely undetectable unless neurons were incubated with XAV (S2 Fig, panels a and b respectively). In contrast with the impact of XAV, overexpression of NEUROD1 promoted significant elevation of NR0B1 but not NR0B2 (Fig 5, panels (b) and (d)). These data show that over a 25-day incubation, the addition of XAV correlates with increased NR0B1/2, but major reductions in both NEUROD1 and A9 phenotype markers. In contrast, an elevation of NEUROD1 results in elevated NR0B1 but also reduced expression of A9 phenotype markers. These data led us to speculate that NR0B1 is a critical regulator of dopaminergic neuronal phenotype so we investigated the idea that NR0B1 could directly impact expression of dopaminergic markers. We overexpressed NR0B1 and NR0B2 in mature midbrain cultures and found that overexpression of NR0B1 significantly upregulated expression of both NR0B1 and NR0B2 (Fig 6) while promoting significant reductions in a number of key dopaminergic neuronal phenotype markers; TH, KCNJ6, ALDH1A1 and LMX1A, but not SLC6A3, SLC18A2 or NR4A2. Curiously TUBB3 was profoundly downregulated by both NR0B1 and NR0B2 overexpression (Fig 6). Consistent with changes in transcript NR0B1 and NR0B2 overexpression increased expression of NR0B1 and NR0B2 (S3 Fig, panels a and b, respectively).
Panels (a) and (b) respectively show the effects of the WNT modulators and NEUROD1 overexpression upon NR0B1. Panels (c) and (d) show effects upon NR0B2 expression. Results are expressed as mean ± SEM of three (biological) replicate experiments. Analysis of panel (a) was via one-way ANOVA with post-hoc Dunnett’s test. For panels (b) and (d) analysis was via Student’s t-tests (at each time point as in Fig 4). For panel (c) analysis was via a non-parametric Kruskal-Wallis test with post-hoc Dunn’s test to accommodate undetectable levels of NR0B2 in two out of three samples, both undetectable values were assigned to 0.0028 (equal to the lowest value detected).
Expression patterns following transfection with NR0B1 (green), NR0B2 (red) or blank (blue) plasmids show that NR0B1 transfection increases both NR0B1 and NR0B2 and impacts expression of NEUROD1, TH, KCNJ6, LMX1A and TUBB3. Control wells underwent transfection with an empty plasmid. Results are expressed as mean ± SEM of three (biological) replicate experiments. Analysis was performed with One-way ANOVA followed by post-hoc Dunnett’s test where each post transfection timepoint represented a repeated measure (t = 0 was not included in analyses). * and ** indicate P<0.05 and P<0.01, respectively. S3 Fig shows increased NR0B1 and NR0B2 immunolabelling up to six days after transfection.
How are NEUROD1 and NR0B1 and NR0B2 impacted in Parkinson’s disease?
We found the idea of a specific regulation of dopaminergic neuron identity fascinating. So we assessed PD array databases, initially looking at the genes defining dopaminergic neuron development and phenotype (such as FOXA2, LMX1A, EN1, EN2, SLC18A2, TH, DDC, ALDH1A1, NR4A2, PITX3, SOX6) as well as more generic markers of a neuronal phenotype (SNCA, SYP, FABP7, GCH1, PTS, TUBB3, MAP2). This analysis showed loss of multiple neuronal transcripts: SNCA, SYP, FABP7, GCH1, PTS, TUBB3, MAP2 (Fig 7, in green) alongside losses of transcripts associated with dopaminergic neuronal function (SLC18A2, TH, ALDH1A1, KCNJ6 Fig 7, in red) and smaller changes in other transcripts associated with a dopaminergic neuronal phenotype, such as EN1, NR4A2, LMO3 and PBX1 (Fig 7, in blue). Curiously, some transcripts were unaffected by PD (FOXA2, LMX1A, PITX3 and NR0B2) while some; MSX1, MSX2, ASCL1, OTX2, and most importantly, NEUROD1 and NR0B1 were all significantly elevated. We used human tissues from control and individuals with PD to check known dopaminergic and midbrain markers protein expression levels and found PITX3/TH immunolabelling (S4 Fig) to be broadly consistent with transcriptomic data.
The initial group (red) shows transcripts whose loss is usually correlated with PD, for example TH and SLC6A3. The middle group (blue) shows how PD impacts upon developmental and mature neuron transcripts. The right-hand grouping (green) shows neuronal markers, rather than dopaminergic neuronal markers. The dotted line indicates the average loss of neuronal markers. The ΔCt values for each control and PD transcript were analysed using Student’s paired t-tests. *, ** = P<0.05, 0.01, respectively (n = 8–10 datasets).
Given the impact that a loss of β-catenin-dependent WNT signalling has upon dopaminergic neuron markers in vitro, we evaluated the PD databases for changes in soluble and membrane bound WNT signalling components. Thus, we looked for changes in frizzled receptors, leucine rich repeat containing G-protein coupled receptors (LGR4, 5 & 6) and LDL receptor related proteins (LRP4, 5 & 6), Kremen1 & Kremen2, ROR1 & ROR2 and RYK. We also looked for changes in soluble WNT signalling regulators (WNT-inhibitory factor (WIF1), secreted frizzled related proteins (SFRP1, 2, 4 & 5), dickkopfs (DKK1, 2, 3 & 4), disheveled (DVL1, DVL2, DVL3) and the R-spondins (RSPO1, 2, 3 & 4). This analysis showed some profound changes, particularly a strong reduction of RSPO2, a gene encoding the secreted protein roof plate-specific spondin-2, an activator of β-catenin-dependent WNT signalling (Fig 8 summarises the statistically significant, PD-associated changes in WNT signalling regulators). All changes in WNT signal-related transcript are shown in S5 Fig. Significantly, we found that overexpression of NR0B1, but not NR0B2 promoted reduction in RSPO2 expression from cultures (Fig 6).
Membrane bound components increased (teal) while soluble factors increased and decreased (magenta). Analysis is identical that in Fig 3.
One particularly interesting element of this study is the finding that the absence of β-catenin-dependent WNT signalling (XAV) promoted a profound elevation of SOX6 transcript expression. This is noteworthy as there are no putative LEF1 binding sites and very few putative NEURO D binding sites regulating transcription (see S5 Table), which may indicate additional mechanisms underlying gene regulation.
Discussion
Transplantation therapies for PD rely upon appropriately patterned dopaminergic neural progenitors to supplement dopamine production from the surviving A9 neurons. Progenitors at or before day 20 of differentiation are used for transplantation since they offer increased connectivity and integration with host neurons as they mature (for reviews see [23, 24]). The use of these cells is predicated on the assumption that following patterning, dopaminergic progenitors do not undergo substantive changes after transplantation. Protocols for creating midbrain neural progenitors, in vitro, draw heavily from our understanding of mouse neuronal differentiation and rely upon sonic hedgehog and WNT signalling to regulate development (see [5, 6, 8, 48]) as defined by expression of transcription factors such as EN1/2, OTX2, FOXA2, LMX1A, MSX1, NEUROG2 and later, ASCL1, NR4A2, PITX3 [5–9] and PBX1 [10]. Within the adult midbrain, dopaminergic neurons of the substantia nigra (A9) and the adjacent ventral tegmental area (A10) are both characterized by expression of TH: markers such as aldehyde dehydrogenase (ALDH1A1, [13]), KCNJ6 (G Protein-Activated Inward Rectifier Potassium Channel 2, [14]), the calcium binding protein, calbindin (CALB1), orthodenticle homeobox 2 (OTX2) and SRY (Sex Determining Region Y)-Box 6 (SOX6, see [15]) are subsequently used to define A9 (ALDH1A1, KCNJ6, SOX6) and A10 (OTX2, CALB1) neurons.
Over the course of 80 days our maturing cultures showed expression of markers consistent with the development and maturation of dopaminergic neurons; early markers such as FOXA2 and LMX1A decrease (but do not disappear) after day 20 of differentiation as markers of a more mature dopaminergic neuronal phenotype such as TH, SLC18A2 and synuclein increase and plateau. These findings are broadly consistent with changes in gene expression in late developing cultures identified previously [36]. However, a key consideration in understanding how developing dopaminergic cultures develop and maintain a phenotype is that, at least in mouse development, early genes/patterning factors do not necessarily “switch off”, but remain expressed in post-mitotic neurons (see [12]). This idea provides the platform for the current study. Recent work has shown that excessive activation of the β-catenin-dependent WNT pathway results in gene expression changes of midbrain floorplate progenitors and decreased numbers of mDA neurons. Interestingly, this work also shows a severe reduction of dopaminergic neurons at the rostral midbrain region (location of A9 neurons), whilst the caudal midbrain regions (location of A10 neurons) were not affected [49]. This finding supports the idea that A9 and A10 neurons may be regulated, in part, by β-catenin-dependent WNT signalling [50]. With this in mind we investigated the impact of CHIR, a commonly used inhibitor of glycogen synthase kinase 3β (GSK3β) which, along with other sonic hedgehog signalling mimetics, is used as a driver of dopaminergic differentiation [35, 51–53]. We also investigated the impact of XAV939, a tankyrase inhibitor that antagonizes β-catenin-dependent WNT signalling [54], which is used drive cortical differentiation [55]. We opted to assess the impact of WNT signalling upon cultures that had achieved relative stability, ie after day 40 when transcripts for TH, LMX1A and TUBB3 had largely stabilized, thus we added WNT signalling modulators from days 40 to 65 of maturation. CHIR reduced expression of two gene transcripts commonly associated with dopaminergic neuron maturation; NR4A2 and TH, indicating a very subtle shift away from a dopaminergic phenotype following 25 days of activation. In comparison, XAV significantly reduced expression of many genes associated with a midbrain dopaminergic neuronal phenotype including SLC18A2, KCNJ6, ALDH1A1, TH, PITX3, LMX1A, OTX2, and EN2. In support of these findings, XAV and to a lesser extent, CHIR, appeared to reduce the intensity of TH, KCNJ6 and SLC6A3 immunolabeling. XAV did not, however, affect markers of a neuronal phenotype (β3-tubulin or synuclein) indicating a shift away from a dopaminergic neuronal phenotype (rather than a neuronal phenotype). That the proposed markers of an A9 phenotype: SOX6 [56], KCNJ6 (GIRK2 [16]) and ALDH1A1 [17] showed different expression patterns following XAV raises questions around the choice of a single marker to identify A9 neurons. The decreased nestin expression in response to XAV might implicate β-catenin-dependent WNT signalling in regulating neural progenitor growth [57], although nestin has also been shown to positively regulate β-catenin-dependent WNT signalling as well [58], suggesting a positive feedback loop. Confounding this is the finding of a XAV-mediated reduction of WNT5A transcript. While Wnt5a plays a role in mouse dopaminergic neuron development [8], this finding may indicate an active regulatory WNT signalling feedback loop present in cultures [59]. As the activation of β-catenin-dependent-WNT signalling with CHIR in our system has little effect, but inhibiting that system has a profound influence, we suggest that cultures possess active β-catenin-dependent-WNT signalling pathways that provide ongoing maintenance of dopaminergic phenotype. While our data shows a dynamic regulation of WNT5A transcript, other WNT signalling regulators such as dickkopf 3 [19] and Rspo2 (r-spondin 2) (86) have been shown to regulate dopaminergic neuron development and phenotype. At present the nature of the ligands contributing to WNT signalling to regulate dopaminergic neuron phenotype in culture is unclear.
One particularly interesting finding was that XAV significantly reduced the expression of NEUROD1; a member of the basic helix-loop-helix (bHLH) family of transcription factors linked to neurogenesis [60]. It is expressed in the adult murine substantia nigra [40] and while overexpression of NeuroD diminishes the capacity of Nr4a2 to drive expression of Th, dopamine transporter, Vmat2 and dopa decarboxylase in rat neural precursor cells [41] it has been used as one of four factors to reprogram astrocytes to dopaminergic neurons [42]. Our data show that NEUROD1 transcript expression is reduced by XAV, which is consistent with evidence that NeuroD1 expression is upregulated by β-catenin-dependent WNT signalling (upon removal of Sox2 mediated transcriptional repression [21]). Given the role of NeuroD1 in reprogramming astrocytes to dopaminergic neurons [42] we surmised that overexpression of NEUROD1 regulate markers of a dopaminergic neuronal phenotype. In contrast to expectations NEUROD1 overexpression reduced markers of an A9 dopaminergic neuron phenotype. That it is specific for a handful of genes, but does not correlate with the distribution of MotifMap identified putative NEUROD transcription factor binding sites indicates a complex relationship between NEUROD1 expression and dopaminergic (and in particular A9) phenotype. The nature of this relationship might be clarified by considering the role that the transcriptional repressors, SHP1 (NR0B2) and DAX (NR0B1) play in regulating gene expression in cultured neurons. While XAV reduced NEUROD1 it profoundly elevated both NR0B1 and NR0B2 expression. NR0B1 and NR0B2 can inhibit expression of not only NEUROD1 regulated genes [43], but also HNF4 regulated genes [45, 46], and MotifMap analysis indicates an abundance of putative HNF4 regulatory elements controlling dopaminergic neuronal phenotype. That both XAV as well as NEUROD1 overexpression promoted elevations of both NR0B1 and NR0B2 may be consistent with a loss of the ability of NEUROD1, and perhaps HNF4, to drive expression of genes related to dopaminergic neuronal phenotype. Couple this with a XAV driven loss of β-catenin-dependent WNT signalling (ie. reduced LEF1 activity) and it is perhaps not surprising that there is a loss of dopaminergic neuronal phenotype in culture. While NEUROD1 overexpression elevates NR0B1 it is worth noting that NR0B1 overexpression drives NR0B2 and NEUROD1 (possibly indicating the presence of a feedback loop). Thus, over expression of NR0B1 and NR0B2 is capable of selectively supressing TH and LMX1A as well as markers of an A9 dopaminergic neuronal phenotype (ALDH1A1 and KCNJ6). Curiously we also show that NR0B1 and NR0B2 overexpression reduces expression of TUBB3 (which has an abundance of putative NEUROD transcription factor binding sites), a finding that may indicate suppression or degradation of axon guidance and/or maintenance.
One curious finding of this study was the XAV-mediated increase in SOX6 expression. Sox2 and Sox6 form a positive feedback loop to regulate neuronal differentiation [61] and decreasing Sox6 expression in 3T3L1 cells increases levels of β-catenin [62]. Exactly how SOX6 is elevated by XAV is unclear, since there are few putative NEUROD1 or LEF1 transcription factor binding sites regulating expression, indeed MotifMap analysis indicates the regulation of SOX6 expression to be quite distinct from the other genes of interest, including NR0B1 and NR0B2.
Given our evidence showing that inhibition of β-catenin-dependent WNT signalling reduced a number of markers of an A9 midbrain dopaminergic phenotype while increasing NR0B1 and NR0B2 expression, we wondered whether WNT signalling systems could be perturbed in PD. We initially assessed PD array databases (GEO accession numbers: GSE20163, GSE20164, GSE20141 [27], GSE20164 [28], GSE20333 (Edna et al., [unpublished]), GSE7307 [29], GDS3128 [30], GDS3129 [30], GSE54282 [31], GSE43490 [32]) to look at changes in genes defining dopaminergic neuron development and phenotype as well as more generic markers of a neuronal phenotype. Consistent with the idea that PD is associated with the loss of dopaminergic neurons, this analysis showed significant and consistent loss of transcripts for multiple neuronal genes: SNCA, SYP, FABP7, GCH1, PTS, TUBB3, MAP2. We also showed very profound and expected loss of transcripts associated with dopaminergic neuronal function (SLC18A2, TH, ALDH1A1, KCNJ6), but smaller losses of other transcripts associated with a dopaminergic neuronal phenotype, such as EN1, NR4A2, LMO3 and PBX1. A most curious finding was that a number of genes associated with a dopaminergic phenotype (for example, FOXA2, EN2, LMX1A and PITX3) did not show reductions in transcript, while some; OTX2, MSX1/2 ASCL1, NEUROD1, SOX2, SOX6, and NR0B1 increased. We then looked for changes in WNT signalling components in the PD array databases, reasoning that changes in WNT signalling transcripts might provide a clue to the regulation of the genes showing such a profound depression in PD (ie. SLC18A2, TH, ALDH1A1, SLC6A3 and KCNJ6). We looked for changes in soluble and membrane bound components of the WNT signalling cascade and found a significant upregulation of some cellular and soluble components of the WNT signalling cascade, most notably a profound loss of RSPO2 transcript. R-spondins are secreted ligands of leucine-rich repeat containing G protein-coupled receptors that enhance WNT signalling [63]. Consistent with the idea of a loss of β-catenin-dependent WNT signalling is a PD associated increase in multiple membrane bound WNT signalling components (RYK, ROR2, LGR5, LGR5, FZD7, FZD9), along with increased WNT1, 2B and 11, as well as WIFI and DKK2. Interestingly DKK3 transcript was also reduced; DKK3 plays a role in midbrain dopaminergic neuron development [19]. Supporting the idea that WNT signalling may be perturbed in PD we identified that both NEUROD1 and NR0B1 are significantly upregulated. Since, in culture, overexpression of NEUROD1 promotes expression of NR0B1 and since NR0B1 overexpression reduces markers of a dopaminergic, and in particular an A9 phenotype, as well as reducing RSPO2 expression, we expect that elevated nigral NEUROD1 and NR0B1 expression would induce the same effect, ie. a reduction in the appearance of A9 markers along with a reduction in RSPO2 expression. This hypothesis clearly needs further investigation.
We believe that our in vitro model data support the idea that at least a proportion of the genes essential for the characterization of dopaminergic phenotype are regulated by β-catenin-dependent-WNT signalling either directly by NEUROD1 expression, or indirectly the elevation of NR0B1 (possibly working in conjunction with NR0B2). Confounding our efforts to identify genes regulated by XAV and or NEUROD1/NR0B1/NR0B2 overexpression is that multiple signalling systems are activated or inhibited by the presence of the neurotrophic factors added to cultures; these include GDNF, BDNF, TGFβ3, dbcAMP as well as the notch inhibitor, DAPT. These factors can regulate expression of many genes identified in this study, including MAFB, PDX1, TEF-1, CREB and ETS2 [64–71]. In addition, while our culture protocol enriches for dopaminergic neurons, both forebrain and hindbrain markers are present in differentiations [36]. Whether and or how these other cell types contribute to WNT signalling is currently unknown.
In summary, we have shown that WNT signalling plays an ongoing role in regulating dopaminergic neuronal phenotype, even late in maturation. As part of this process the loss of β-catenin-dependent WNT signalling in cultures profoundly increases expression of the transcriptional repressors, NR0B1/2 which in-turn downregulate dopaminergic, and in particular A9 neuronal markers, as well as RSPO2. The PD array database profiling indicates that tissue of the substantia nigra has reduced RSPO2 and elevated NR0B1 (and NEUROD1) leading us to believe that dysregulation of β-catenin-dependent WNT signalling is evident in the substantia nigra of individuals with PD. While this dysregulation may exacerbate the degeneration-induced loss of dopaminergic neurons we believe that, regardless of the site of transplantation, local brain chemistry will have a profound impact upon the ability of transplanted neural progenitors to maintain a dopaminergic neuronal phenotype.
Supporting information
S2 Table. Taqman probes used for quantitative PCR.
https://doi.org/10.1371/journal.pone.0261730.s002
(DOCX)
S4 Table. Primary and secondary antibodies used in this study.
https://doi.org/10.1371/journal.pone.0261730.s004
(DOCX)
S5 Table. MotifMap analysis of putative transcriptional control.
Genes of interest are listed at the top of each column. Pairs of columns show transcription factor and putative number of binding sites. Sites were estimated using MotifMap (http://motifmap.ics.uci.edu/). Note: Putative binding sites are not corrected for overlap or orientation.
https://doi.org/10.1371/journal.pone.0261730.s005
(XLSX)
S1 Fig. Post-NEUROD1 transfection immunolabelling.
Panels (a)-(e) show Hoechst nuclear labelling, LMX1A-eGFP, NEUROD1, β3-tubulin and colour combined images seven days after NEUROD1 transfection. Panels (f)-(j) show Hoechst nuclear labelling, LMX1A-eGFP, NEUROD1, β3-tubulin and colour combined images seven days after control transfection. Images are all 20x (scale bar is 100μm) taken with the same imaging parameters. The insets show x4 images.
https://doi.org/10.1371/journal.pone.0261730.s006
(TIF)
S2 Fig. NR0B1 and NR0B2 expression following incubation of midbrain cultures with XAV.
From left to right the top pair of panels show DAPI nuclear labelling, LMX1A-eGFP, NR0B1, β3-tubulin and colour combined images after 14 days of vehicle (control, upper) or XAV (100nM, lower). The bottom pair of panels show identical labelling except where NR0B2 is indicated. Images are all 20x (scale bar is 100μm) taken with the same imaging parameters.
https://doi.org/10.1371/journal.pone.0261730.s007
(TIF)
S3 Fig. Transfection with NR0B1 and NR0B2 increases protein expression.
From left to right the panels DAPI nuclear labelling, LMX1A-eGFP, NR0B1 (top three panels) & NR0B2 (bottom three panels) and colour combined images in vehicle (control, 6 days) and at 3 and 6 days after transfection. Images are all 20x (scale bar is 100μm) taken with the same imaging parameters across each of the days.
https://doi.org/10.1371/journal.pone.0261730.s008
(TIF)
S4 Fig. Typical PD and control immunolabelling.
Panels show tyrosine hydroxylase (green, panels a and e) and PITX3 (red, panels b and f) in sections of substantia nigra from an age-matched control (panels a-d) and a PD patient (panels e-h). Panels (c) and (g) show Hoechst labelled nuclei, while panels (d) and (h) show colour combined images. Note the widespread reduction in green fluorescence in PD. Scale bar indicates 25μm.
https://doi.org/10.1371/journal.pone.0261730.s009
(TIF)
S5 Fig. PD-related changes in WNT signalling components.
Panel (a) shows changes in nigral WNTs, panel (b) shows changes in soluble WNT signalling ligands while panels (c) and (d) show changes in frizzled receptors and cellular WNT-regulators or signal transduction components. Although data is shown as a ratio PD/control, the ΔCt values for matched control and PD transcript arrays analysed using Student’s paired t-tests. *, ** = P<0.05, 0.01, respectively (n = 6–10 datasets).
https://doi.org/10.1371/journal.pone.0261730.s010
(TIF)
Acknowledgments
The Victorian Brain Bank is supported by The Florey Institute of Neuroscience and Mental Health, The Alfred and the Victorian Institute of Forensic Medicine and funded in part by Parkinson’s Victoria, MND Victoria and Cure for MND Foundation
References
- 1. Thomas B, Beal MF. Parkinson’s disease. Hum Mol Genet. 2007;16 Spec No. 2:R183–94. pmid:17911161.
- 2. Klockgether T. Parkinson’s disease: clinical aspects. Cell Tissue Res. 2004;318(1):115–20. pmid:15365814.
- 3. Chaudhuri KR, Martinez-Martin P, Schapira AH, Stocchi F, Sethi K, Odin P, et al. International multicenter pilot study of the first comprehensive self-completed nonmotor symptoms questionnaire for Parkinson’s disease: the NMSQuest study. Mov Disord. 2006;21(7):916–23. pmid:16547944.
- 4. Allen SJ, Watson JJ, Shoemark DK, Barua NU, Patel NK. GDNF, NGF and BDNF as therapeutic options for neurodegeneration. Pharmacol Ther. 2013;138(2):155–75. pmid:23348013.
- 5. Prakash N, Brodski C, Naserke T, Puelles E, Gogoi R, Hall A, et al. A Wnt1-regulated genetic network controls the identity and fate of midbrain-dopaminergic progenitors in vivo. Development. 2006;133(1):89–98. Epub 2005/12/13. pmid:16339193.
- 6. Tang M, Villaescusa JC, Luo SX, Guitarte C, Lei S, Miyamoto Y, et al. Interactions of Wnt/{beta}-Catenin Signaling and Sonic Hedgehog Regulate the Neurogenesis of Ventral Midbrain Dopamine Neurons. J Neurosci. 2010;30(27):9280–91. pmid:20610763
- 7. Joksimovic M, Anderegg A, Roy A, Campochiaro L, Yun B, Kittappa R, et al. Spatiotemporally separable Shh domains in the midbrain define distinct dopaminergic progenitor pools. Proc Natl Acad Sci U S A. 2009;106(45):19185–90. pmid:19850875.
- 8. Andersson ER, Salto C, Villaescusa JC, Cajanek L, Yang S, Bryjova L, et al. Wnt5a cooperates with canonical Wnts to generate midbrain dopaminergic neurons in vivo and in stem cells. Proc Natl Acad Sci U S A. 2013;110(7):E602–10. pmid:23324743; PubMed Central PMCID: PMC3574960.
- 9. Toledo EM, Gyllborg D, Arenas E. Translation of WNT developmental programs into stem cell replacement strategies for the treatment of Parkinson’s disease. Br J Pharmacol. 2017;174(24):4716–24. pmid:28547771; PubMed Central PMCID: PMC5727333.
- 10. Smidt MP, Smits SM, Burbach JP. Homeobox gene Pitx3 and its role in the development of dopamine neurons of the substantia nigra. Cell Tissue Res. 2004;318(1):35–43. pmid:15300495.
- 11. Hwang DY, Hong S, Jeong JW, Choi S, Kim H, Kim J, et al. Vesicular monoamine transporter 2 and dopamine transporter are molecular targets of Pitx3 in the ventral midbrain dopamine neurons. J Neurochem. 2009;111(5):1202–12. pmid:19780901; PubMed Central PMCID: PMC4896488.
- 12. Arenas E, Denham M, Villaescusa JC. How to make a midbrain dopaminergic neuron. Development. 2015;142(11):1918–36. pmid:26015536.
- 13. Smidt MP, Burbach JP. How to make a mesodiencephalic dopaminergic neuron. Nat Rev Neurosci. 2007;8(1):21–32. pmid:17180160.
- 14. Li J, Dani JA, Le W. The role of transcription factor Pitx3 in dopamine neuron development and Parkinson’s disease. Curr Top Med Chem. 2009;9(10):855–9. Epub 2009/09/17. pmid:19754401; PubMed Central PMCID: PMC2872921.
- 15. Luk KC, Rymar VV, van den Munckhof P, Nicolau S, Steriade C, Bifsha P, et al. The transcription factor Pitx3 is expressed selectively in midbrain dopaminergic neurons susceptible to neurodegenerative stress. J Neurochem. 2013;125(6):932–43. pmid:23331067.
- 16. Thompson L, Barraud P, Andersson E, Kirik D, Bjorklund A. Identification of dopaminergic neurons of nigral and ventral tegmental area subtypes in grafts of fetal ventral mesencephalon based on cell morphology, protein expression, and efferent projections. J Neurosci. 2005;25(27):6467–77. Epub 2005/07/08. 25/27/6467 [pii]. pmid:16000637.
- 17. Liu G, Yu J, Ding J, Xie C, Sun L, Rudenko I, et al. Aldehyde dehydrogenase 1 defines and protects a nigrostriatal dopaminergic neuron subpopulation. J Clin Invest. 2014;124(7):3032–46. pmid:24865427; PubMed Central PMCID: PMC4071380.
- 18. Fu Y, Paxinos G, Watson C, Halliday GM. The substantia nigra and ventral tegmental dopaminergic neurons from development to degeneration. J Chem Neuroanat. 2016;76(Pt B):98–107. pmid:26859066.
- 19. Fukusumi Y, Meier F, Gotz S, Matheus F, Irmler M, Beckervordersandforth R, et al. Dickkopf 3 Promotes the Differentiation of a Rostrolateral Midbrain Dopaminergic Neuronal Subset In Vivo and from Pluripotent Stem Cells In Vitro in the Mouse. J Neurosci. 2015;35(39):13385–401. pmid:26424886.
- 20. Sancho RM, Law BM, Harvey K. Mutations in the LRRK2 Roc-COR tandem domain link Parkinson’s disease to Wnt signalling pathways. Hum Mol Genet. 2009;18(20):3955–68. Epub 2009/07/25. pmid:19625296; PubMed Central PMCID: PMC2748899.
- 21. Kuwabara T, Hsieh J, Muotri A, Yeo G, Warashina M, Lie DC, et al. Wnt-mediated activation of NeuroD1 and retro-elements during adult neurogenesis. Nat Neurosci. 2009;12(9):1097–105. Epub 2009/08/25. pmid:19701198; PubMed Central PMCID: PMC2764260.
- 22. Zhang J, Gotz S, Vogt Weisenhorn DM, Simeone A, Wurst W, Prakash N. A WNT1-regulated developmental gene cascade prevents dopaminergic neurodegeneration in adult En1(+/-) mice. Neurobiol Dis. 2015;82:32–45. pmid:26049140.
- 23. Arenas E. Towards stem cell replacement therapies for Parkinson’s disease. Biochem Biophys Res Commun. 2010;396(1):152–6. pmid:20494130.
- 24. Lindvall O, Kokaia Z. Prospects of stem cell therapy for replacing dopamine neurons in Parkinson’s disease. Trends Pharmacol Sci. 2009;30(5):260–7. pmid:19362379.
- 25. Gennet N, Tamburini C, Nan X, Li M. FolR1: a novel cell surface marker for isolating midbrain dopamine neural progenitors and nascent dopamine neurons. Sci Rep. 2016;6:32488. Epub 2016/09/02. pmid:27580818; PubMed Central PMCID: PMC5007511.
- 26. Fathi A, Mirzaei M, Dolatyar B, Sharifitabar M, Bayat M, Shahbazi E, et al. Discovery of Novel Cell Surface Markers for Purification of Embryonic Dopamine Progenitors for Transplantation in Parkinson’s Disease Animal Models. Mol Cell Proteomics. 2018;17(9):1670–84. Epub 2018/06/01. pmid:29848781; PubMed Central PMCID: PMC6126395.
- 27. Zheng B, Liao Z, Locascio JJ, Lesniak KA, Roderick SS, Watt ML, et al. PGC-1alpha, a potential therapeutic target for early intervention in Parkinson’s disease. Sci Transl Med. 2010;2(52):52ra73. pmid:20926834; PubMed Central PMCID: PMC3129986.
- 28. Lesnick TG, Papapetropoulos S, Mash DC, Ffrench-Mullen J, Shehadeh L, de Andrade M, et al. A genomic pathway approach to a complex disease: axon guidance and Parkinson disease. PLoS Genet. 2007;3(6):e98. pmid:17571925; PubMed Central PMCID: PMC1904362.
- 29. Roth RB, Hevezi P, Lee J, Willhite D, Lechner SM, Foster AC, et al. Gene expression analyses reveal molecular relationships among 20 regions of the human CNS. Neurogenetics. 2006;7(2):67–80. pmid:16572319.
- 30. Duke DC, Moran LB, Pearce RK, Graeber MB. The medial and lateral substantia nigra in Parkinson’s disease: mRNA profiles associated with higher brain tissue vulnerability. Neurogenetics. 2007;8(2):83–94. pmid:17211632.
- 31. Riley BE, Gardai SJ, Emig-Agius D, Bessarabova M, Ivliev AE, Schule B, et al. Systems-based analyses of brain regions functionally impacted in Parkinson’s disease reveals underlying causal mechanisms. PLoS One. 2014;9(8):e102909. pmid:25170892; PubMed Central PMCID: PMC4149353.
- 32. Corradini BR, Iamashita P, Tampellini E, Farfel JM, Grinberg LT, Moreira-Filho CA. Complex network-driven view of genomic mechanisms underlying Parkinson’s disease: analyses in dorsal motor vagal nucleus, locus coeruleus, and substantia nigra. Biomed Res Int. 2014;2014:543673. pmid:25525598; PubMed Central PMCID: PMC4261556.
- 33. Moran LB, Duke DC, Deprez M, Dexter DT, Pearce RK, Graeber MB. Whole genome expression profiling of the medial and lateral substantia nigra in Parkinson’s disease. Neurogenetics. 2006;7(1):1–11. Epub 2005/12/14. pmid:16344956.
- 34. Niclis JC, Gantner CW, Alsanie WF, McDougall SJ, Bye CR, Elefanty AG, et al. Efficiently Specified Ventral Midbrain Dopamine Neurons From Human Pluripotent Stem Cells Under Xeno-Free Conditions Restore Motor Deficits in Parkinsonian Rodents. Stem Cells Transl Med. 2016. pmid:28297587.
- 35. Kriks S, Shim JW, Piao J, Ganat YM, Wakeman DR, Xie Z, et al. Dopamine neurons derived from human ES cells efficiently engraft in animal models of Parkinson’s disease. Nature. 2011;480(7378):547–51. Epub 2011/11/08. nature10648 [pii]. pmid:22056989.
- 36. Watmuff B, Hartley BJ, Hunt CP, Fabb SA, Pouton CW, Haynes JM. Human pluripotent stem cell derived midbrain PITX3(eGFP/w) neurons: a versatile tool for pharmacological screening and neurodegenerative modeling. Front Cell Neurosci. 2015;9(104):1–15. pmid:25873861; PubMed Central PMCID: PMC4379917.
- 37. Daily K, Patel VR, Rigor P, Xie X, Baldi P. MotifMap: integrative genome-wide maps of regulatory motif sites for model species. BMC Bioinformatics. 2011;12:495. pmid:22208852; PubMed Central PMCID: PMC3293935.
- 38. Kreft L, Soete A, Hulpiau P, Botzki A, Saeys Y, De Bleser P. ConTra v3: a tool to identify transcription factor binding sites across species, update 2017. Nucleic Acids Res. 2017;45(W1):W490–W4. Epub 2017/05/05. pmid:28472390; PubMed Central PMCID: PMC5570180.
- 39. Lee C, Huang CH. LASAGNA-Search: an integrated web tool for transcription factor binding site search and visualization. Biotechniques. 2013;54(3):141–53. Epub 2013/04/20. pmid:23599922.
- 40. Mesman S, Kruse SJ, Smidt MP. Expression analyzes of early factors in midbrain differentiation programs. Gene Expr Patterns. 2018;27:8–15. Epub 2017/09/30. pmid:28958789.
- 41. Park CH, Kang JS, Kim JS, Chung S, Koh JY, Yoon EH, et al. Differential actions of the proneural genes encoding Mash1 and neurogenins in Nurr1-induced dopamine neuron differentiation. J Cell Sci. 2006;119(Pt 11):2310–20. Epub 2006/05/26. pmid:16723737.
- 42. Rivetti di Val Cervo P, Romanov RA, Spigolon G, Masini D, Martin-Montanez E, Toledo EM, et al. Induction of functional dopamine neurons from human astrocytes in vitro and mouse astrocytes in a Parkinson’s disease model. Nat Biotechnol. 2017;35(5):444–52. Epub 2017/04/12. pmid:28398344.
- 43. Kim JY, Chu K, Kim HJ, Seong HA, Park KC, Sanyal S, et al. Orphan nuclear receptor small heterodimer partner, a novel corepressor for a basic helix-loop-helix transcription factor BETA2/neuroD. Mol Endocrinol. 2004;18(4):776–90. Epub 2004/01/31. pmid:14752053.
- 44. Iyer AK, Zhang YH, McCabe ER. LXXLL motifs and AF-2 domain mediate SHP (NR0B2) homodimerization and DAX1 (NR0B1)-DAX1A heterodimerization. Mol Genet Metab. 2007;92(1–2):151–9. Epub 2007/08/10. pmid:17686645; PubMed Central PMCID: PMC2065763.
- 45. Conti B, Porcu C, Viscomi C, Minutolo A, Costantini S, Corazzari M, et al. Small heterodimer partner 1 directly interacts with NS5A viral protein and has a key role in HCV related liver cell transformation. Oncotarget. 2016;7(51):84575–86. Epub 2016/09/24. pmid:27661118; PubMed Central PMCID: PMC5356682.
- 46. Nedumaran B, Hong S, Xie YB, Kim YH, Seo WY, Lee MW, et al. DAX-1 acts as a novel corepressor of orphan nuclear receptor HNF4alpha and negatively regulates gluconeogenic enzyme gene expression. J Biol Chem. 2009;284(40):27511–23. Epub 2009/08/05. pmid:19651776; PubMed Central PMCID: PMC2785680.
- 47. Santiago JA, Potashkin JA. Network-based metaanalysis identifies HNF4A and PTBP1 as longitudinally dynamic biomarkers for Parkinson’s disease. Proc Natl Acad Sci U S A. 2015;112(7):2257–62. Epub 2015/02/04. pmid:25646437; PubMed Central PMCID: PMC4343174.
- 48. Joksimovic M, Yun BA, Kittappa R, Anderegg AM, Chang WW, Taketo MM, et al. Wnt antagonism of Shh facilitates midbrain floor plate neurogenesis. Nat Neurosci. 2009;12(2):125–31. pmid:19122665.
- 49. Nouri N, Patel MJ, Joksimovic M, Poulin JF, Anderegg A, Taketo MM, et al. Excessive Wnt/beta-catenin signaling promotes midbrain floor plate neurogenesis, but results in vacillating dopamine progenitors. Mol Cell Neurosci. 2015;68:131–42. pmid:26164566; PubMed Central PMCID: PMC4633300.
- 50. Smits SM, Burbach JP, Smidt MP. Developmental origin and fate of meso-diencephalic dopamine neurons. Prog Neurobiol. 2006;78(1):1–16. pmid:16414173.
- 51. Nolbrant S, Heuer A, Parmar M, Kirkeby A. Generation of high-purity human ventral midbrain dopaminergic progenitors for in vitro maturation and intracerebral transplantation. Nat Protoc. 2017;12(9):1962–79. pmid:28858290.
- 52. Kirkeby A, Grealish S, Wolf DA, Nelander J, Wood J, Lundblad M, et al. Generation of regionally specified neural progenitors and functional neurons from human embryonic stem cells under defined conditions. Cell Rep. 2012;1(6):703–14. pmid:22813745.
- 53. Denham M, Bye C, Leung J, Conley BJ, Thompson LH, Dottori M. Glycogen synthase kinase 3beta and activin/nodal inhibition in human embryonic stem cells induces a pre-neuroepithelial state that is required for specification to a floor plate cell lineage. Stem Cells. 2012;30(11):2400–11. pmid:22911885; PubMed Central PMCID: PMC3533765.
- 54. Huang SM, Mishina YM, Liu S, Cheung A, Stegmeier F, Michaud GA, et al. Tankyrase inhibition stabilizes axin and antagonizes Wnt signalling. Nature. 2009;461(7264):614–20. Epub 2009/09/18. pmid:19759537.
- 55. Qi Y, Zhang XJ, Renier N, Wu Z, Atkin T, Sun Z, et al. Combined small-molecule inhibition accelerates the derivation of functional cortical neurons from human pluripotent stem cells. Nat Biotechnol. 2017;35(2):154–63. pmid:28112759.
- 56. Panman L, Papathanou M, Laguna A, Oosterveen T, Volakakis N, Acampora D, et al. Sox6 and Otx2 control the specification of substantia nigra and ventral tegmental area dopamine neurons. Cell Rep. 2014;8(4):1018–25. pmid:25127144.
- 57. Yue Y, Xue Q, Yang J, Li X, Mi Z, Zhao G, et al. Wnt-activated olfactory ensheathing cells stimulate neural stem cell proliferation and neuronal differentiation. Brain Res. 2020;1735:146726. Epub 2020/02/15. pmid:32057808.
- 58. Zhao Z, Lu P, Zhang H, Xu H, Gao N, Li M, et al. Nestin positively regulates the Wnt/beta-catenin pathway and the proliferation, survival and invasiveness of breast cancer stem cells. Breast Cancer Res. 2014;16(4):408. Epub 2014/07/25. pmid:25056574; PubMed Central PMCID: PMC4220087.
- 59. Bisson JA, Mills B, Paul Helt JC, Zwaka TP, Cohen ED. Wnt5a and Wnt11 inhibit the canonical Wnt pathway and promote cardiac progenitor development via the Caspase-dependent degradation of AKT. Dev Biol. 2015;398(1):80–96. Epub 2014/12/09. pmid:25482987.
- 60. Sharma A, Moore M, Marcora E, Lee JE, Qiu Y, Samaras S, et al. The NeuroD1/BETA2 sequences essential for insulin gene transcription colocalize with those necessary for neurogenesis and p300/CREB binding protein binding. Mol Cell Biol. 1999;19(1):704–13. Epub 1998/12/22. PubMed Central PMCID: PMC83927. pmid:9858593
- 61. Lee KE, Seo J, Shin J, Ji EH, Roh J, Kim JY, et al. Positive feedback loop between Sox2 and Sox6 inhibits neuronal differentiation in the developing central nervous system. Proc Natl Acad Sci U S A. 2014;111(7):2794–9. Epub 2014/02/07. pmid:24501124; PubMed Central PMCID: PMC3932859.
- 62. Leow SC, Poschmann J, Too PG, Yin J, Joseph R, McFarlane C, et al. The transcription factor SOX6 contributes to the developmental origins of obesity by promoting adipogenesis. Development. 2016;143(6):950–61. Epub 2016/02/20. pmid:26893351.
- 63. Lebensohn AM, Rohatgi R. R-spondins can potentiate WNT signaling without LGRs. Elife. 2018;7. Epub 2018/02/07. pmid:29405118; PubMed Central PMCID: PMC5800842.
- 64. Kim H. The transcription factor MafB promotes anti-inflammatory M2 polarization and cholesterol efflux in macrophages. Sci Rep. 2017;7(1):7591. Epub 2017/08/10. pmid:28790455; PubMed Central PMCID: PMC5548719.
- 65. Zhu Y, Liu Q, Zhou Z, Ikeda Y. PDX1, Neurogenin-3, and MAFA: critical transcription regulators for beta cell development and regeneration. Stem Cell Res Ther. 2017;8(1):240. Epub 2017/11/04. pmid:29096722; PubMed Central PMCID: PMC5667467.
- 66. Yao X, Li K, Liang C, Zhou Z, Wang J, Wang S, et al. Tectorigenin enhances PDX1 expression and protects pancreatic beta-cells by activating ERK and reducing ER stress. J Biol Chem. 2020;295(37):12975–92. Epub 2020/07/22. pmid:32690606; PubMed Central PMCID: PMC7489913.
- 67. Dey A, Varelas X, Guan KL. Targeting the Hippo pathway in cancer, fibrosis, wound healing and regenerative medicine. Nat Rev Drug Discov. 2020;19(7):480–94. Epub 2020/06/20. pmid:32555376; PubMed Central PMCID: PMC7880238.
- 68. Ambrosino C, Iwata T, Scafoglio C, Mallardo M, Klein R, Nebreda AR. TEF-1 and C/EBPbeta are major p38alpha MAPK-regulated transcription factors in proliferating cardiomyocytes. Biochem J. 2006;396(1):163–72. Epub 2006/02/24. pmid:16492136; PubMed Central PMCID: PMC1449985.
- 69. Yordy JS, Muise-Helmericks RC. Signal transduction and the Ets family of transcription factors. Oncogene. 2000;19(55):6503–13. Epub 2001/02/15. pmid:11175366.
- 70. Kramer ER, Liss B. GDNF-Ret signaling in midbrain dopaminergic neurons and its implication for Parkinson disease. FEBS Lett. 2015;589(24 Pt A):3760–72. pmid:26555190.
- 71. Yeung HY, Chan DK, Mak NK, Wagner GF, Wong CK. Identification of signal transduction pathways that modulate dibutyryl cyclic adenosine monophosphate activation of stanniocalcin gene expression in neuroblastoma cells. Endocrinology. 2003;144(10):4446–52. Epub 2003/09/10. pmid:12959990.