Altered glutamate response and calcium dynamics in iPSC-derived striatal neurons from XDP patients

Experimental Neurology 308 (2018) 47–58

Contents lists available at ScienceDirect

Experimental Neurology journal homepage: www.elsevier.com/locate/yexnr

Research Paper

Altered glutamate response and calcium dynamics in iPSC-derived striatal neurons from XDP patients

T

⁎

P. Capetiana,b, ,1, N. Stanslowskyc,1, E. Bernhardia, K. Grütza, A. Domingoa, N. Brüggemanna,b, ⁎ M. Naujockc, P. Seiblera, C. Kleina, ,2, F. Wegnerc,2 a

Institute of Neurogenetics, University of Lübeck, Germany Department of Neurology, University of Lübeck, Germany c Department of Neurology, Hannover Medical School, Germany b

A R T I C LE I N FO

A B S T R A C T

Keywords: XDP iPSC Spiny projection neurons Electrophysiology AMPA receptor Calcium dynamics

X-linked dystonia-parkinsonism (XDP) is a neurodegenerative disorder endemic to Panay Island (Philippines). Patients present with generalizing dystonia and parkinsonism. Genetic changes surrounding the TAF1 (TATAbox binding protein associated factor 1) gene have been associated with XDP inducing a degeneration of striatal spiny projection neurons. There is little knowledge about the pathophysiology of this disorder. Our objective was to generate and analyze an in-vitro model of XDP based on striatal neurons differentiated from induced pluripotent stem cells (iPSC). We generated iPSC from patient and healthy control fibroblasts (3 affected, 3 controls), followed by directed differentiation of the cultures towards striatal neurons. Cells underwent characterization of immunophenotype as well as neuronal function, glutamate receptor properties and calcium dynamics by whole-cell patch-clamp recordings and calcium imaging. Furthermore, we evaluated expression levels of AMPA receptor subunits and voltage-gated calcium channels by quantitative real-time PCR. We observed no differences in basic electrophysiological properties. Application of the AMPA antagonist NBQX led to a more pronounced reduction of postsynaptic currents in XDP neurons. There was a higher expression of AMPA receptor subunits in patient-derived neurons. Basal calcium levels were lower in neurons derived from XDP patients and cells with spontaneous calcium transients were more frequent. Our data suggest altered glutamate response and calcium dynamics in striatal XDP neurons.

1. Introduction X-linked dystonia-parkinsonism (XDP, DYT/PARK-TAF1, or DYT3 in the old nomenclature) is a rare neurodegenerative disorder initially restricted to the Island of Panay (Visayas, Philippines). Extremely rare on a global scale, the prevalence in the Philippines is 0.31 per 100,000 and as high as 5.74 per 100,000 on Panay Island (Lee et al., 2011). Thus, it represents a considerable challenge to the local healthcare system in this circumscribed region. The disease is characterized by an initial dystonic stage, followed by combined dystonia-parkinsonism, and eventually, a parkinsonian phase (Rosales, 2010). Histopathological studies proposed an initial loss of striatal GABAergic spiny projection neurons (SPN) in the striosomal compartment, followed by a degeneration of SPN throughout the entire striatum (Goto et al., 2005). The reason for this focal degeneration is unclear. The disease locus has

⁎

been mapped to a ~ 427 kbp region on Xq13.1 (Domingo et al., 2015) Although no protein-coding genetic variation could be identified, five disease-specific single nucleotide changes (DSC1, DSC2, DSC3, DSC10, DSC12), one 48-bp deletion, and one 2.6-kbp retrotransposon insertion segregate in affected individuals (Domingo et al., 2015; Nolte et al., 2003). All changes lie either within intronic regions, intergenic DNA segments, or in a proposed non-conventional exon of the neighboring TAF1 gene (TATA-box binding protein associated factor 1). Downregulation of TAF1 and retention of intron 32 (Aneichyk) have been proposed as disease mechanism (Herzfeld et al., 2013; Sako et al., 2011; Makino et al., 2007). However, it is unknown how reduced expression levels of this ubiquitously expressed gene lead to a tissue-specific degeneration of SPN. Recent data strongly support the pathogenicity of the SVA retrotransposon insertion in TAF1 (Aneichyk et al., 2018; Rakovic et al., 2018). However, the complexity of the genetic variations

Corresponding authors at: Institute of Neurogenetics, University of Lübeck, Ratzeburger Allee 160, 23562 Lübeck, Germany. E-mail addresses: [email protected] (P. Capetian), [email protected] (C. Klein). Both authors contributed equally 2 Both authors contributed equally 1

https://doi.org/10.1016/j.expneurol.2018.06.012 Received 26 February 2018; Received in revised form 26 May 2018; Accepted 21 June 2018 Available online 23 June 2018 0014-4886/ © 2018 Elsevier Inc. All rights reserved.

Experimental Neurology 308 (2018) 47–58

P. Capetian et al.

employed antibodies see supplementary Table 2). RNA was extracted from all iPSC clones as well as spontaneously differentiated embryoid bodies (EBs), transcribed into cDNA and analyzed by quantitative realtime PCR (please see below). Expression levels of iPSC lines were compared to fibroblasts for the pluripotency markers NANOG, GDF3, OCT4, and SOX2 and relative values calculated. EBs, generated as published (Xu et al., 2001), were analyzed for the expression of germ layer markers, i.e. endoderm (AFP, GATA4, SOX17), mesoderm (RUNX1, MSX1, MYH6), and ectoderm (NCAM, PAX6). Relative expression levels of these values were calculated by comparing them to the levels of the initial iPSC lines. Evaluation of genomic stability of each iPSC line in comparison to the parental fibroblast line was accomplished by SNP analysis with the Infinium OmniExpress-24 beadchip (Illumina) according to the manufacturer's protocol. DNA was extracted from all iPSC clones and compared to the DNA of the respective fibroblast line. Retrieved data was analyzed using the Karyo Studio software (Illumina).

in XDP and their uncertain effects complicate the generation of appropriate animal models. Induced pluripotent stem cells (iPSC) are reprogrammed from patient specific somatic cells (Takahashi et al., 2007). Due to their pluripotency, the directed differentiation to the desired cell types (including striatal neurons) is possible (Carri et al., 2013; Stanslowsky et al., 2016a). The patient-specific neurons harbor all genetic changes inherent to the original cell type reprogrammed. Therefore, this approach seems particularly feasible for disease modeling when the causative genetic changes are unknown or unclear. At present, there is no study available about the mechanisms leading to the neuronal degeneration in XDP. We sought to analyze striatal neurons derived by directed differentiation from human iPSC lines. Three XDP patients and three ethnically- and age-matched healthy volunteers served as donors. We focused on functional analyses using whole-cell patch-clamp as well as calcium imaging with a special emphasis on the glutamatergic transmission and calcium dynamics. In addition, we measured expression levels of AMPA-receptor subunits and voltage-gated calcium channels (CaV) by quantitative real-time polymerase chain reaction (qPCR).

2.2. Directed differentiation of iPSC lines to striatal neurons For the striatal differentiation of iPSC, we adapted an already established protocol recently published elsewhere (Stanslowsky et al., 2016a): After manual removal of areas suspicious of spontaneous differentiation, iPSC colonies were incubated with collagenase IV (Gibco) for 10 min at 37 °C. Colonies were removed from the culture dish, triturated carefully to smaller pieces and plated on ultra-low attachment culture plates (Cornig) for the formation of free-floating aggregates known as EB. EBs were kept for two days in knockout serum replacement (KSR) medium with the addition of the following small molecules: The effectivity of neural induction was enhanced by 100 nM LDN193189 (Stemgent) and 10 μM SB-431542 (Tocris) from day 0–6 (Chambers et al., 2009). The Rho kinase inhibitor Y-27632 (Stemcell Technologies) was added at a concentration of 10 μM for the enhancement of cell survival during the first two days (Ungrin et al., 2008). Rostral patterning was enhanced by the addition of 1 μM of the antagonist at the WNT-pathway IWP-2 (Stemcell Technologies) from day 0–10, while 0.2 μM of the smoothened agonist Purmorphamine from day 4–10 enhanced ventral patterning (Li et al., 2009). Medium was changed to KSR/N2 medium 1:1 on day two. From day 4 onwards, EBs were kept in N2-medium until day 12. EBs were reduced in size by careful trituration and plated on 2.5% Matrigel (MG, Cornig) coated 12well plates in N2B27 medium with BDNF 20 ng/ml (PeproTech), GDNF 10 ng/ml (PeproTech) and db-cAMP 50 μM (Enzo Life Sciences). Each time when reaching full confluency (~ day 20 and day 40), cells were removed mechanically from the culture dish, triturated to smaller clumps and replated on matrigel (MG) coated culture ware. For terminal differentiation, aggregates of neural cells were plated on poly-Dlysine (Sigma Aldrich) and laminin (Roche) coated coverslips (Hecht). For enhanced survival, at each passaging step, 10 μM of Y-27632 were added to the medium. Medium was changed every other day. Cells were analyzed after 90 days. For an overview of the protocol, see Fig. 2a. Detailed media composition is listed in supplementary Table 1. We performed a total of two differentiations of each iPSC line (3 cell lines each for XDP and controls).

2. Methods 2.1. Generation and cultivation of human iPSC lines Fibroblast cultures from three genetically proven and clinically affected XDP patients in their dystonic phase and three matched healthy volunteers from the Philippines (see Table 1 for control and patient details), who gave informed written consent according to the ethical regulations of the University of Lübeck, were included in this study (AZ12–219). All parts of this project were approved by the ethical committee of the University of Lübeck and were conducted in full accordance to their regulations. Genotyping of patients, collection of skin biopsies and generation of fibroblast cultures were performed in a former study from our group (Domingo et al., 2016). Cells were kept in basic fibroblast medium (see supplementary Table 1 for an overview of media composition) in T-300 flasks (Cornig) and passaged by TrypLE (Thermo Fisher Scientific). Fibroblasts were reprogrammed integration-free by commercially available Sendai virus vectors according to the manufacturer's guidelines (CytoTune®-iPS Reprogramming Kit, Invitrogen). We picked emerging iPSC colonies manually and cultivated them on irradiated mouse embryonic fibroblasts (MTI-GlobalStem) in iPSC medium with 10 ng/μl of freshly added fibroblast growth factor 2 (FGF2, Millipore). Established iPSC lines were characterized by immunofluorescent stainings for pluripotent markers (for an overview of Table 1 Characteristics of control subjects and XDP patients as skin fibroblast donors in this study. Controls: ID code

Sex

Age at biopsy

L8357 L8360 L8361

m m m

37 35 35

2.3. Immunofluorescent staining and microscopy Immunofluorescent stainings were performed at three time points: 1. After the first replating step at day 20 in order to evaluate early transcription factors associated with ventral forebrain identity. 2. After day 90 to quantify the amount of terminally differentiated SPN. 3. After whole-cell patch-clamp recordings to quantify the amount of correctly identified SPN (see below). Cells on coverslips were fixed for 15 min in 4.5% freshly prepared paraformaldehyde (PFA, Sigma-Aldrich). After washing cells 3× for 15 min with phosphate-buffered saline (PBS, Gibco), unspecific

XDP patients: ID code

Sex

Age at biopsy

Age at onset

Disease duration (y)

Disease stage

L-5748 L-7994 L-7149

m m m

44 37 51

38 33 48

6 4 3

parkinsonian > dystonic parkinsonian > dystonic dystonic > parkinsonian

48

Experimental Neurology 308 (2018) 47–58

P. Capetian et al.

field from two coverslips each of two independent differentiations per iPSC line. Between 12 and 24 cells per coverslip (depending on the cell number and density in one high power field of view) were assayed.

epitopes were blocked by incubation with blocking buffer containing 5% of normal serum from the corresponding host species (SigmaAldrich) of the secondary antibody and 0,1% Triton X-100 (SigmaAldrich) in PBS. Afterwards, primary antibodies (see supplementary Table 2) were diluted in incubation buffer containing 1% of normal serum from the corresponding host species and 0,1% Triton X-100 in PBS and incubated with the cells overnight at 4 °C. The following day, cells were washed 3× with washing buffer (0,1% Triton X-100 in PBS. Next, cells were incubated with the respective secondary antibodies diluted in incubation buffer for 2 h at room temperature. Following another washing step (3× washing buffer), auto-fluorescence was quenched by incubating cells in 0.3% SudanB (Roth) in 70% ethanol for 5 min. After a further washing step (15 min washing buffer and 5 min PBS), coverslips were mounted onto slides (Menzel Gläser) with VECTASHIELD containing 1.5% DAPI (Biozol). In order to determine background staining, primary antibodies were omitted for one additional coverslip per differentiation. Biocytine/DARPP-32 stainings were performed similarly, except for the fact that biocytine was visualized by fluorochrome-coupled streptavidin instead of a secondary antibody. Specimens were visualized with a LSM 710 confocal laser scanning microscope (Zeiss) and the ZEN Black software. All images were acquired with a 40× oil-immersion lens (Zeiss). For cell counting, in a first step cell background was determined by scanning specimens stained without the first antibody. Then, six random high-power fields were taken and counted by manually with ImageJ (NIH). Image acquisition and counting were carried out observer-blinded. We counted one coverslip per staining from two independent differentiations of every line.

2.6. Quantitative real-time PCR For RNA extraction cells were collected and processed with the RNeasy kit (Qiagen) according to the manufacturer's directions. A total amount of 500 ng of RNA was transcribed into cDNA with the Maxima First Strand cDNA Synthesis Kit (Thermo Fisher Scientific) according to the standard protocol. Quantitative real-time PCR for pluripotency and germ layer associated genes was performed as follows: A total amount of 25 ng of RNA transcribed to cDNA was analyzed per reaction. QPCR reaction and analysis was carried out with the LC SYBR Green Mix (Fermentas) in a Lightcycler 480 (Roche) under the following conditions: 95 °C for 5 min, followed by 45 cycles of 95 °C for 10 s, 60 °C for 10 s and 72 °C for 10 s. For the reactions 1.75 μM of forward and reverse primers were used. The expression of all marker genes was normalized to the expression of ß-actin and calculated by the relative standard curve method. Mean values of two technical replicates per line were calculated. Standard curves were generated by serial dilutions of original cDNA. Quantitative real-time PCRs for CaV channels and AMPA receptor subunits were performed according to the following protocol: cDNA from 15 ng total RNA per reaction was analyzed with Power SYBRGreen PCR Master Mix (Life Technologies) in a StepOnePlus instrument (Applied Biosystems). The amplification conditions were as follows: 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. For the reactions 1.75 μM of forward and reverse primers were used. Threshold cycle (Ct) values of the targets were normalized against the endogenous reference β2-microglobulin (Ct (target) – Ct (reference) = ΔCt). Two technical replicates of two independent differentiations per iPSC line were analyzed. For a list of primer sequences, see supplementary Table 3.

2.4. Electrophysiology Cells were plated on coated glass coverslips and differentiated as described above. Two coverslips each of two independent differentiations per line were employed. For electrophysiological measurements the coverslips were transferred to 35 mm culture dishes (Greiner) filled with external solution containing 142 mM NaCl, 8 mM KCl, 1 mM CaCl2, 10 mM glucose and 10 mM HEPES (Gibco), adjusted to pH 7.4 with NaOH (325 mOsm) (all reagents from Sigma-Aldrich). Mg2+ was omitted in the external solution in order to deblock NMDA receptors. The internal pipette solution consisted of 153 mM KCl, 1 mM MgCl2, 10 mM HEPES, 5 mM EGTA and 2 mM Mg-ATP, adjusted to pH 7.3 with KOH (305 mOsm) (all reagents from Sigma-Aldrich). During some recordings 0.1% biocytin was included in the pipette solution to identify the phenotype of the neuron. Whole-cell patch-clamp recordings were performed as reported previously (Stanslowsky et al., 2016b). Memantine (MEM, 100 μM) 2.3-dihydroxy-6-nitro-7-sulphamoyl-benzo(f) quinoxaline (NBQX, 10 μM), or bicuculline (BIC, 10 μM) (all reagents from Sigma-Aldrich) were applied via gravity flow using a modified SF77B perfusion fast-step system (Warner Instruments).

2.7. Statistics All statistic data was analyzed with GraphPad Prism 5 (GraphPad Software). Results of cell counting and qPCR data were first averaged per iPSC line, then we used the means per line in the statistic tests mentioned below. For the relative amounts of immunofluorescent markers and expression levels of AMPA receptors and CaV means were calculated per line, pooled for controls and patients and the two groups compared by two-way analysis of variance (ANOVA) and a Bonferroni post-test. Immunofluorescent stainings were also analyzed individually to test for significant differences between lines (two-way ANOVA with Tukey correction for multiple analyzes). For the relative amount of cells with PSC, means were calculated per line, pooled for controls and patients and the two groups compared by the non-parametric MannWhitney test. PSC amplitudes and frequencies, were analyzed by first calculating means of measured neurons from control and XDP lines and then performing an unpaired t-tests with correction for multiple testing following the Holm-Sidak method. Sodium (INa) and potassium (IK) current amplitudes at different holding potentials were compared by two-way analysis of variance (ANOVA) and Bonferroni post-test. Basic neurophysiological properties (Table 2) were analyzed by first calculating means of control and XDP lines and then performing a nonparametric Mann-Whitney test. Spiking behavior, AP-amplitudes and mPSC frequency block data was analyzed by the non-parametric MannWhitney test after calculating means from control and XDP neurons. Calcium imaging data were compared using either unpaired t-tests (relative amount of cells with spontaneous calcium release, mean values of lines pooled for analysis) or the Mann-Whitney test in case of non-Gaussian distribution (basal calcium content, response amplitudes und relative amount of responsive cells, cells pooled for controls and

2.5. Calcium Imaging Intracellular calcium transients were measured in Fura 2-AM-loaded cells (Sigma-Aldrich) with the Till Vision Imaging System (T.I.L.L. Photonics) as described previously (Stanslowsky et al., 2014). With the increase of cytosolic free calcium, the 340/380 nm excitation ratio of Fura 2-AM increases, which allows to monitor spontaneous or provoked intracellular Ca2+ changes upon separate application of the glutamate receptor agonists a-amino-3-hydroxy-5-methyl-4-isoxazolepropionic (AMPA, 100 μM), N-methyl-D-aspartate (NMDA, 100 μM), or glutamate (50 μM). Cells were imaged under baseline conditions for 70–90 s. Afterwards, the agonist was applied and cells were imaged for another 90–130 s. Measurements were terminated by application of the depolarizing agent potassium chloride (KCl, 50 mM) to ensure viability and excitability of the recorded cells. For statistical analysis, only Ca2+ rises of R340/380 ≥ 0.05 were considered. We analyzed one high power 49

Experimental Neurology 308 (2018) 47–58

P. Capetian et al.

Table 2 Functional properties of iPSC-derived striatal neurons from XDP patients and healthy controls. Whole-cell patch-clamp recordings measured voltage-gated sodium (INa) and potassium (IK) currents, resting membrane potentials (RMP), cell capacitances (C membrane), action potential (AP) and postsynaptic current (PSC) properties. Data are given as means ± s.e.m. per individual cell line. There were no significant differences between the means of controls and XDP (p > 0.05, nonparametric Mann-Whitney test). Control Functional properties INa max. Amplitudes (pA/pF) IK max. Amplitudes (pA/pF) RMP (mV) C membrane (pF) Input resistance (GOhm) Cells with single APs (%) Cells with repetitive APs (%) AP amplitudes (mV) Cells with spontaneous APs (%) Cells with PSCs (%) PSC frequencies (Hz) PSC amplitudes (pA)

8361–2 (n = 40) −126.8 ± 15.5 123.3 ± 10.6 −36.2 ± 1.5 13.0 ± 0.9 1.25 ± 0.20 48 20 40.9 ± 4.0 12 86.1 0.91 ± 0.19 52.8 ± 1.7

XDP 8357–2 (n = 14) −109.0 ± 16.4 117.1 ± 8.4 −36.8 ± 1.8 14.0 ± 1.7 1.19 ± 0.28 63 7 46.9 ± 3.6 10 85.7 1.28 ± 0.29 38.9 ± 1.3

8360–5 (n = 28) −56.0 ± 8.0 83.1 ± 11.7 −33.7 ± 2.1 23.6 ± 1.4 0.47 ± 0.05 74.6 6.2 35.6 ± 2.9 7.7 82.1 0.34 ± 0.08 46.5 ± 3.1

5748–2 (n = 14) −120.4 ± 31.1 127.2 ± 32.3 −37.8 ± 2.0 19.0 ± 2.0 0.90 ± 0.09 41.6 8.4 41.3 ± 4.8 0 93.8 0.57 ± 016 53.7 ± 2.6

7994–1 (n = 21) −105.8 ± 15.7 96.0 ± 12.4 −39.8 ± 2.5 18.4 ± 2.4 2.44 ± 0.55 58.3 33.4 47.1 ± 3.1 8.3 92.3 1.02 ± 0.29 27.5 ± 1.2

7149–2 (n = 12) −118.1 ± 22.7 145.3 ± 17.8 −32.6 ± 2.1 19.6 ± 1.5 1.48 ± 0.60 75 25 41.7 ± 2.8 33.3 91.7 0.63 ± 0.12 57.4 ± 3.4

cells co-expressed DARPP-32. Approximately, 30% of all DARPP-32positive cells co-expressed CTIP2, a transcription factor required for fully differentiated SPN (Arlotta et al., 2008) (Fig. 2d, f). There were no obvious differences in respect to the immunophenotype between individual iPSC lines as well as neurons derived from healthy controls and XDP patients (Fig. 2e, f).

XDP before analysis). All data are presented as means ± s.e.m. Significance level was set as p < 0.05. 3. Results 3.1. Successful generation of iPSC lines by Sendai virus (SeV) transduction 20 days after SeV transduction, colonies with a characteristic morphology became apparent and reached a sufficient size for selection. After manual picking and propagation, established lines were characterized for the pluripotency markers Nanog, SSEA4, TRA-1-60 and Oct4 (Fig. 1a, b). All lines showed efficient upregulation of transcription factors related to pluripotency in comparison to the fibroblast line (Fig. 1c). Pluripotency and differentiation potential were additionally analyzed by quantitative real-time PCR of different germ layer markers. Each embryoid body stage cell line had elevated levels of at least two markers of each germ layer (Fig. 1d). We found no significant differences between the iPSC lines in terms of gene expression. As described before (Hussein et al., 2011), most iPSC lines acquired few and small new genetic changes during the reprogramming process in comparison to the source fibroblast lines. Larger insertions, deletions or translocations could be excluded (see supplementary Table 4 for a detailed overview of the observed genetic changes).

3.3. Elevated AMPA-mediated glutamatergic innervation of XDP neurons To assess voltage-gated ion channel and action potential properties as well as synaptic functionality, we performed electrophysiological recordings of neurons derived from XDP patients and healthy controls. There was no method available to label neurons for markers associated with SPN before recordings. However, we chose medium-sized multipolar neurons with oval soma (a morphology we could associate with DARPP-32-positive neurons in the fluorescent stainings). As a post-hoc quality control, some cells exemplarily received intracellular filling with biocytin during recordings, which was included in the pipette solution, allowing co-staining with anti-DARPP-32 after recordings to ensure predominant SPN identity. In total, 75% of the recorded cells were DARPP-32-positive (we identified 12 biocytine positive cells after recording, of which 9 stained positive for DARPP-32, Fig. 3a). All cells in the patient and control group showed potassium outward and sodium inward currents upon stepwise incremental depolarization (Fig. 3b). Potassium and sodium current amplitudes normalized to individual cell capacitances did not significantly differ in both groups (Fig. 3c). There were no significant differences for basic electrophysiological properties between control and XDP cells (see Table 2 for means ± s.e.m. and sample sizes). The vast majority of all cells investigated fired at least single action potentials (APs) after depolarizing current injections (XDP 58.3 ± 9.6%, Ctrl 61.9 ± 7.7%, Fig. 3d, e). Repetitive firing was present only in a minor population in both the XDP and control group with no significant differences in percentage (XDP 22.3 ± 7.3%, Ctrl 11.1 ± 4.5%, Fig. 3e) or AP amplitudes (XDP 43.7 ± 1.9 mV, Ctrl 39.3 ± 2.1 mV, Fig. 3f). Spontaneous spiking was observed in 13.9 ± 10.0% of XDP cells and 9.9 ± 1.2% of controls. Neither the percentage of cells receiving postsynaptic input nor the frequency or amplitudes of postsynaptic currents (PSCs) were significantly different in both groups (Fig. 4a). By blocking PSCs with different antagonists of ligand-gated ion channels, we determined that the predominant type of synaptic input to the striatal neurons was GABAergic. By application of bicuculline (BIC), a GABAA receptor blocker, PSC frequencies were reduced to 28.2 ± 10.6% in XDP and 20.5 ± 5.5% in control neurons (Fig. 4b, c). The AMPA receptor antagonist NBQX inhibited PSCs significantly only in XDP cells, suggesting that XDP neurons may receive an elevated AMPA-mediated

3.2. Directed differentiation towards striatal neurons 20 days after the start of differentiation from iPSC towards neurons, the majority of cells expressed transcription factors associated with forebrain development and specification of the ventral telencephalon as the primordial region of the striatum: FOXG1 (expressed by 71,3% of all control cells and 73% of all XDP cells), a mandatory transcription factor for forebrain development (Martynoga et al., 2005) and OTX2 (expressed by 85,4% of all control cells and 84,7% of all XDP cells), involved in the specification of the forebrain and midbrain (Boyl et al., 2001) (Fig. 2b, e). The transcription factor MEIS2 (expressed by 78,9% of all control cells and 86,3% of all XDP cells) is associated with the subventricular zone (SVZ) of the lateral ganglionic eminence (LGE) as the origin of striatal neurons (Toresson et al., 2000a) and FOXP1 (expressed by 80,1% of all control cells and 91,1% of all XDP cells) is a transcription factor of striatal neuronal precursors before final specification as SPN (Precious et al., 2016) (Fig. 2c, e). After 90 days of maturation, roughly 80% of all cells displayed a neuronal phenotype with expression of the microtubule-associated protein 2 (MAP2, Fig. 2d, f). Of all cells, 30% expressed the dopamine- and cAMP-regulated neuronal phosphoprotein 32 (DARPP-32) that is associated with SPN identity (Anderson and Reiner, 1991), while 40% of all MAP2-positive 50

Experimental Neurology 308 (2018) 47–58

P. Capetian et al.

(caption on next page)

51

Experimental Neurology 308 (2018) 47–58

P. Capetian et al.

Fig. 1. Immunofluorescent staining of iPSC colonies after reprogramming. Proliferating colonies of all induced pluripotent stem cell (iPSC) lines expressed factors associated with pluripotency, such as (a) SSEA4 (green) and Nanog (red), as well as (b) TRA-1-60 (green) and OCT4 (red) in contrast to the surrounding feeder cells. Nuclear staining was done with 4′,6-Diamidin-2-phenylindol (DAPI, blue), scale bars = 50 μm. Quantitative real-time PCR showed a robust increase of pluripotency genes in relation to the original fibroblasts for all iPSC lines without significant differences between control and XDP-lines (c). After spontaneous differentiation into embryoid bodies (EB), cell lines displayed an upregulation of genes associated with all three germ layers in comparison to the original iPSC without significant differences between control and XDP-lines (d). (2way ANOVA + Bonferroni post-test). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

et al., 2016). A recent study employed the same cell types to explore the correlation of a hexameric repeat within the SVA (SINE/VNTR/Alu) retrotransposon in intron 32 on disease onset (Bragg et al., 2017). Yet another study used iPSC, neural stem cells, and GABAergic neurons derived from iPSC as well as induced cortical neurons (iN) from XDP patients and controls to explore the impact of the mentioned SVA retrotransposon on TAF1 expression and performed expression profiling by RNAseq30. The last study employed a different maturation protocol for deriving GABAergic neurons reporting a similar efficacy in generating DARPP32-positive neurons (Arber et al., 2015). Another very recent study from our group employed the same cohort of patient derived cells in conjunction with the same maturation protocol as well as cortical differentiation (Arber et al., 2015). We could demonstrate that the removal of the SVA retrotransposon by CRISPR/Cas9 normalized TAF1 expression levels in iPSC but to a lesser extent in mature neurons. All these studies employed patient derived cells in a bottom-up approach connecting the genetic changes to transcriptional changes. We followed a top-down approach from the differentiated neuronal subtype most strongly affected from XDP-pathology and characterized its aberrant behavior in a neuronal network in vitro. These approaches complement each other and future studies will hopefully close the gap between these two as well as in our understanding of XDP pathology. Protocols for directed differentiation of iPSC towards striatal neurons have been available for a few years (Carri et al., 2013; Aubry et al., 2008) and have successfully been employed for modeling neurodegenerative diseases such as choreoacanthocytosis (ChAc) (Stanslowsky et al., 2016a). By using a similar protocol, we succeeded in generating striatal neurons both from XDP patients and healthy controls in vitro. A particular challenge for all studies dealing with striatal differentiation protocols is the number of medium SPN present after differentiation. While SPN represent a vast majority in the adult striatum (> 95% of all neurons) (Gerfen, 1992), directed differentiation of iPSC leads to SPN of roughly 40% of all neurons in vitro (Carri et al., 2013; Aubry et al., 2008), a percentage we could not surpass in our study. One potential reason might be “contaminating” precursors of the medial ganglionic eminence (MGE) where GABAergic interneurons of the cortex and striatum but no medium SPN arise (Kelsom and Lu, 2013). However, even a recent protocol for differentiation of iPSC towards SPN that used activin instead of the more classical SHH agonist/WNT antagonist approach for patterning could not achieve > 40% DARPP-32-positive neurons, although suppressing the expression of MGE-associated genes (Arber et al., 2015). We would therefore rather suspect insufficient neurotrophic support or maturation in vitro as a potential reason for these shortcomings. While we were quite successful in identifying individual SPN for patch clamp recordings based on cellular morphology (75% of all neurons filled with biocytin could be identified post-hoc as DARPP-32-positive SPN), calcium imaging and qPCR are a necessary analysis for larger bulks of cells. SPN-specific reporter constructs might overcome these shortcomings, but have not yet been established. Despite the rather long time span of differentiation, a small fraction of measured neurons still displayed immature properties like missing ability to generate spikes upon depolarization and absent synaptic activity. The low abundance of astrocytes in differentiating iPSC derived neural cultures has been considered a reason for this as co-cultivation with this cell-type leads to an increase of neuronal maturity (Kuijlaars et al., 2016). However, neither adding primary astrocytes not harboring the genetic alterations associated with XDP, nor deriving astrocytes

glutamatergic input compared to controls. An NMDA-mediated glutamatergic input, indicated by PSC-blockade after application of the NMDA blocker memantine (MEM), could not be observed in either group. 3.4. Altered basal intracellular Ca2+ and spontaneous Ca2+ signals in XDP neurons We found that XDP neurons had significantly reduced basal intracellular Ca2+ levels compared to controls (control 0.44 ± 0.007, XDP 0.40 ± 0.006, p < 0.05; Fig. 5b). Spontaneous Ca2+ transients occurred more often in XDP neurons (control 11.5 ± 2.9%, XDP 27.3 ± 5.4%, p < 0.05; Fig. 5c). The intracellular Ca2+ increase upon application of the glutamate receptor agonists AMPA, NMDA, and glutamate indicated that iPSCderived striatal neurons functionally express ligand-gated AMPA, NMDA and other glutamate receptors (Fig. 5d, e, f). Using calcium imaging, no significant differences between control and XDP cells were observed with respect to the amplitudes of Ca2+ transients and percentage of responding cells (Fig. 5e, f). 3.5. Indication of altered expression of AMPA subunits in neurons from XDP patients AMPA receptor expression in XDP neurons was higher than in controls (Fig. 6a) without reaching statistical significance (p˃0.05). AMPA receptors are composed of four types of subunits: GluR1, GluR2, GluR3, and GluR4, all of which showed a higher expression in XDP, predominantly GluR1, 3 and 4. Genomic expression levels of voltagegated Ca2+ channel subunits were similar in XDP and control neurons (Fig. 6b) and could be excluded as a possible reason for the observed differences in basal Ca2+ concentrations or spontaneous Ca2+ transients. Row factors for GluR (p = 0.1991) and calcium channel (p = 0.1441) were not significant and therefore no significant main effect of the disease. 4. Discussion Neuropathological studies identified striatal projection neurons (SPN) as prone to the neurodegenerative process of XDP and provided a potential explanation of the two-staged clinical course due to a preferential loss of SPN in the striosomal compartment during the early disease phase (Goto et al., 2005). More recent studies showed a preferential striosomal loss of neuropeptide Y (NPY)-positive interneurons demonstrating a more widespread striatal neurodegenerative process than previously thought (Goto et al., 2013). Furthermore, dysbalances of GABAergic innervation in the GPi and GPe were found in XDP-patients by intraoperative microdialysis during deep brain stimulation (Tronnier et al., 2015). However, at the moment it remains unknown why striatal neurons are particularly susceptible to the disease process. A study employing cDNA from primary skin fibroblasts showed changes in transcript levels of > 300 genes in XDP patient cells compared to controls (Domingo et al., 2016). However, as the disease process exclusively affects the central nervous system, in-vitro models based on human neural cells would be preferable. A previous study successfully generated iPSC lines from XDP patients and performed a neural induction leading to proliferative neural stem cells (NSC) (Ito 52

Experimental Neurology 308 (2018) 47–58

P. Capetian et al.

Fig. 2. Immunofluorescent staining during striatal differentiation of iPSC. After 20 days of the directed differentiation protocol (a), the majority of cells were positive for transcription factors associated with (b) forebrain development, such as FOXG1 (green) and OTX2 (red) or (c) striatal commitment, such as FOXP1 (green) and MEIS2 (red). (d) After 90 days of differentiation, cells expressed the mature neuronal marker microtubule-associated protein 2 (MAP2), of which ~40% also contained the dopamine- and cAMP-regulated neuronal phosphoprotein (DARPP-32), associated with striatal medium spiny projection neurons (SPN). Roughly 30% of all DARPP-32-positive cells co-expressed CTIP2, a transcription factor required for terminal differentiation of SPN. We observed no statistically significant differences in the immunophenotype of control and XDP lines for the precursor as well as the mature markers (e, f). Cell nuclei were stained with 4′,6-Diamidin-2-phenylindol (DAPI, blue). Data are presented as means ± s.e.m. (p > 0.05, 2way ANOVA + Bonferroni post-test), scale bars = 20 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

53

Experimental Neurology 308 (2018) 47–58

P. Capetian et al.

Fig. 3. Voltage-gated ion currents and action potential recordings of iPSC-derived striatal neurons from XDP patients and healthy controls. (a) Intracellular filling of neurons during recordings with biocytin (red after streptavidin staining) allowed the post-hoc identification of the immunophenotype. Of all recorded cells, 75% coexpressed the dopamine- and cAMP-regulated neuronal phosphoprotein (DARPP-32, green) associated with striatal medium spiny projection neurons (SPN). Scale bar = 10 μm. (b) Voltage-gated sodium inward and potassium outward currents were recorded in whole-cell voltage-clamp mode by increasing depolarizing steps of 10 mV from a holding potential of −70 to 40 mV. Note, this trace was recorded from an XDP neuron. (c) Ion currents normalized for cell sizes based on the capacitance of the cell membrane (pA/pF) were not significantly different between patient-derived (n = 41) and control cells (n = 82) (p > 0.05, 2way ANOVA + Bonferroni post-test). (d) Example of XDP neurons firing single and repetitive action potentials upon depolarization in current-clamp mode. (e) The percentage of cells with no, single or repetitive action potentials did not differ in both groups (cell numbers for each group included in graph) (p > 0.05, nonparametric Mann-Whitney test). (f) Action potential amplitudes were similar in XDP (n = 29) cells and controls (n = 45) (p > 0.05, non-parametric Mann-Whitney test). Data are presented as means ± s.e.m. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The detailed mechanisms leading to striatal neurodegeneration in XDP are unknown, as only few pathophysiological studies have been performed to date (Domingo et al., 2016; Ito et al., 2016; Vaine et al., 2016). However, since Huntingon's disease (HD) as one of the most extensively studied neurodegenerative diseases involves a pronounced affection of SPN (Vonsattel et al., 1985), it appears likely that there may be similarities to the disease process of XDP. Yet, many of the described cellular dysfunctions attributable to mutant Huntingtin, such as mitochondrial dysfunction (Chen, 2011; Costa and Scorrano, 2012; Smith et al., 2014), do not readily explain the accentuated vulnerability of SPN. One of the few hypotheses that do so, is associated with the extensive glutamatergic innervation of SPN by cortical projections sensitizing these cells to excitotoxicity and cell death in Huntington's disease (Fan and Raymond, 2007; Graham et al., 2009). We therefore considered the neuronal activity and glutamatergic innervation as potentially being altered in XDP. Since intracellular calcium dynamics have been demonstrated as an important link between increased neural activity and resulting cellular damage in neurodegenerative diseases (Zündorf and Reiser, 2011; Imamura et al., 2016; Goldberg et al., 2009), we included calcium imaging in our analyses. Furthermore, we characterized neurons from XDP patients and controls based on changes

from the corresponding iPSC lines by time-consuming protocols (Zhou et al., 2015) appeared to be feasible approaches. NMDA receptor-mediated effects (especially when located extrasynaptically) are also considered an important factor for striatal neurotoxicity (Fan and Raymond, 2007; Kaufman et al., 2012). However, synaptic input through NMDA receptors in our cultures was extremely scarce and precluded reliable conclusions. The most likely reason for this is the fact that striatal neurons in vivo receive their glutamatergic input from cortical projection neurons. This cell type originates, in contrast to striatal neurons, from the dorsal parts of the telencephalon (Toresson et al., 2000b). However, directed differentiation to striatal neurons requires inducing a ventral regional fate in vitro (Li et al., 2009). As we were able to demonstrate, the majority of neurons in our cultures displayed a ventral telencephalon phenotype during differentiation. The low abundance of glutamatergic input is therefore most likely the result of a low number of this cell type in our cultures. The influence of primary cortical projections on striatal neurons in co-culture lead to increased glutamatergic input and maturity of spines (Segal et al., 2003). Co-culture experiments for iPSC-derived neurons are still lacking, but should be considered a promising approach for improving the synaptic glutamatergic input of striatal neurons in vitro. 54

Experimental Neurology 308 (2018) 47–58

P. Capetian et al.

Fig. 4. Spontaneous synaptic activity of XDP and control striatal neurons. (a) The percentages of cells with PSCs, PSC amplitudes and frequencies were not significantly different between XDP patient (n = 41) and healthy control cells (n = 78; p > 0.05, non-parametric Mann-Whitney test). (b) Traces of postsynaptic currents (PSCs) were recorded in whole-cell voltage-clamp mode under control conditions (top) and following application of the AMPA receptor blocker NBQX, the NMDA receptor antagonist memantine (MEM) or the GABAA receptor blocker bicuculline (BIC). Note, these traces were recorded from an XDP neuron (p < 0.05, non-parametric Mann-Whitney test). (c) PSC frequencies were significantly reduced by the application of the GABAA inhibitor BIC in XDP (n = 13) and control cells (n = 12). The AMPA receptor antagonist NBQX showed a significant block of PSCs in XDP cells only (p < 0.05, non-parametric Mann-Whitney test).

signal transduction through AMPA receptors on SPN as well as changes in the relation of expression levels of GluRs (Rocher et al., 2016). The authors interpreted their findings as possible compensatory effects of long-term changes in synaptic structure of SPN. Of course, this might also be true for the observed changes in AMPA signaling in XDP since we did not evaluate synaptic morphology (which would not prove very reliable outside primary cell or in vivo models). However, as our cells represent rather a short term model of the disease, we would consider observed changes as being associated to rather acute effects. We found a higher percentage of neurons from XDP patients exhibiting spontaneous intracellular calcium transients. Dysregulation and mishandling of intracellular calcium has been considered an important component of neuronal damage in a wide variety of neurodegenerative diseases (Zündorf and Reiser, 2011). We would interpret this finding in terms of a potentially increased excitability of XDP neurons. In addition, we found a lowered basal calcium level in XDP neurons compared to controls. Since elevated intracellular calcium levels have been reported as being detrimental for human neurons in fronto-temporal dementia (Imamura et al., 2016), lowered levels could actually be protective. Therefore, compensatory mechanisms might be responsible for this phenomenon. We quantified the expression levels of voltagegated calcium channels (CaV) without finding any differences between XDP and controls. Thus, mere differences in CaV expression do not seem to be responsible for the changes of intracellular calcium load in striatal neurons derived from XDP patients. In general, the differences between XDP and control neurons seemed rather subtle. It might be possible that the application of stressors (e.g. respiratory chain inhibitors, inducers of reactive oxygen species or nutrient deprivation) might be necessary to override cellular compensatory mechanisms and display the full phenotype. Identifying a

in expression levels of AMPA glutamate receptors and voltage-gated calcium channels. The majority of recorded neurons received synaptic input as demonstrated by the presence of PSCs with no differences in amplitude or frequency between XDP and control cells. As to be expected of a predominant striatal phenotype, the most pronounced inhibition of PSC was achieved by application of the GABAA receptor antagonist bicuculline, since most striatal neurons are GABAergic. However, no differences between XDP and control neurons in relation to GABAergic synaptic activity were observed, although one hypothesis of the pathophysiology of dystonia assumes a striatal GABAergic imbalance as causative (Gittis and Kreitzer, 2012), which is caused in XDP most likely through the loss of GABAergic projections. While PSCs in control neurons were unaffected by glutamate receptor antagonists, addition of the AMPA receptor antagonist NBQX achieved a significant reduction of synaptic activity only in patientderived neurons. With an increased AMPA receptor subunit expression in XDP cells (although not reaching significance), we would interpret these finding as possible signs for an increased glutamatergic transmission through AMPA receptors. Although calcium imaging did not show a significant difference in response amplitudes after the application of AMPA, which might be due to lesser specificity for SPN, expression levels of the glutamate receptor subunits GluR3, GluR4 and to a lesser extent GluR1 were elevated in XDP neurons. These three subunits play an important role in neural damage and neurodegeneration (Jayakar and Dikshit, 2004). In contrast to GluR2, they are permeable for divalent cations (particularly Ca2+) and suspected to mediate excitotoxicity in a variety of neurodegenerative diseases (Hollmann et al., 1991; Gécz et al., 1999; Rembach et al., 2004). Studies in a rodent model of HD could also demonstrate increased 55

Experimental Neurology 308 (2018) 47–58

P. Capetian et al.

Fig. 5. Ca2+ dynamics of iPSC-derived striatal neurons from XDP patients and healthy controls. (a) Representative image during Fura-2 based Ca2+ recording of control neurons. (b) Basal intracellular Ca2+ levels were significantly higher in healthy controls (n = 173) compared to XDP neurons (n = 243; p < 0.0001, nonparametric Mann-Whitney test). (c) Significantly more XDP cells displayed spontaneous Ca2+ transients (p < 0.05, unpaired t-test). Upon separate application of the neurotransmitters AMPA, NMDA, and glutamate the intracellular Ca2+ concentration increased as exemplarily illustrated in (d) for the application of AMPA in two neurons of control cell line 8357–2. (e) The percentage of control and XDP cells responding with rapid Ca2+ rises to the separate application of AMPA, NMDA, glutamate, and KCl as well as the intracellular Ca2+ amplitudes in (d) did not significantly differ (p < 0.05, non-parametric Mann-Whitney test). Data are presented as means ± s.e.m.

Consent for publication

distinct cell population (e.g. SPN) in heterogeneous iPSC-derived neural cultures is a very demanding and still unreliable task. In conjunction with pronounced differences between individual iPSC lines (most likely due to the complex experimental protocols where small variations can easily add up), disease-specific findings might be covered by “background noise”. Potential solutions might be specific reporter lines, allowing identification of distinct neural populations (Durieux et al., 2011) and automated cultivation platforms (Marx et al., 2013), reducing inconsistencies introduced by manual cell culture work. In conclusion, our results demonstrate the feasibility of reprogramming fibroblasts taken from XDP patients and healthy controls to iPSC and the successful differentiation towards functional striatal neurons, of which roughly 40% showed an immunophenotype corresponding to SPN. XDP-derived striatal neurons displayed an increased NBQX-inhibition of glutamatergic input by AMPA receptors and a trend to an increased expression of a subset of calcium-sensitive AMPA receptor subunits (GluR1, 3, 4), which may suggest a detrimental effect of AMPA receptor-mediated excitotoxicity in XDP. Neurons derived from XDP patients also exhibited signs of altered calcium dynamics with a higher percentage of cells showing spontaneous calcium transients and an overall lowered intracellular calcium level. These new functional data may have implications for the development of therapeutic strategies in XDP. Supplementary data to this article can be found online at https:// doi.org/10.1016/j.expneurol.2018.06.012.

Written informed consent was obtained from the patients and participants for publication of their individual details in this manuscript. The consent form is held by the authors and is available for review by the Editor-in-Chief. Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on request. Competing interests The authors declare no competing financial interests. Funding CK is the recipient of a career development award from the Hermann and Lilly Schilling Foundation. This study was supported by the Collaborative Center for X-linked Dystonia Parkinsonism. Author's contributions PC: Concept and design, data acquisition, analysis and interpretation (striatal neurons), data interpretation (iPSC, electrophysiology, calcium imaging, expression analysis), manuscript writing. Corresponding author. NS: Concept and design, data acquisition and analysis (electrophysiology, expression analysis), data interpretation (electrophysiology, calcium imaging, expression analysis), manuscript writing.

Ethics approval and consent to participate Patients and healthy volunteers from the Philippines gave informed written consent according to the ethical regulations of the University of Lübeck (AZ12–219).

56

Experimental Neurology 308 (2018) 47–58

P. Capetian et al.

Fig. 6. Expression analysis of striatal neurons from healthy controls and XDP patients by quantitative real-time PCR: (a) Genomic expression of the four AMPA receptor subtypes, GluR1, GluR2, GluR3, and GluR4 was elevated in XDP neurons without reaching statistical significance, whereas differential expression of voltagegated Ca2+ channel subunits could not be observed (b). Data are presented as means ± s.e.m. (n = 3 XDP and control cell lines, p > 0.05, 2way ANOVA + Bonferroni post-test).

References

EB: Data acquisition and analysis (striatal neurons, calcium imaging). Data acquisition (electrophysiology). KG: Concept and design, data acquisition, analysis and interpretation (iPSC), manuscript writing. AD: Acquisition and examination of patients and healthy volunteers, manuscript writing. NB: Acquisition and examination of patients and healthy volunteers, manuscript writing. MN: Data acquisition, analysis and interpretation (calcium imaging). PS: Concept and design, data analysis and interpretation (iPSC). CK: Concept and design, manuscript writing, co-corresponding author. FW: Concept and design, data acquisition, analysis and interpretation (electrophysiology), data interpretation (electrophysiology, calcium imaging, expression analysis), manuscript writing.

Anderson, K.D., Reiner, A., 1991. Immunohistochemical localization of DARPP-32 in striatal projection neurons and striatal interneurons: implications for the localization of D1-like dopamine receptors on different types of striatal neurons. Brain Res. 568, 235–243. Aneichyk, T., et al., 2018. Dissecting the causal mechanism of X-linked dystonia-parkinsonism by integrating genome and transcriptome assembly. Cell 172, 897–909 e21. Arber, C., et al., 2015. Activin A directs striatal projection neuron differentiation of human pluripotent stem cells. Dev. Camb. Engl. 142, 1375–1386. Arlotta, P., Molyneaux, B.J., Jabaudon, D., Yoshida, Y., Macklis, J.D., 2008. Ctip2 controls the differentiation of medium spiny neurons and the establishment of the cellular architecture of the striatum. J. Neurosci. 28, 622–632. Aubry, L., et al., 2008. Striatal progenitors derived from human ES cells mature into DARPP32 neurons in vitro and in quinolinic acid-lesioned rats. Proc. Natl. Acad. Sci. U. S. A. 105, 16707–16712. Boyl, P.P., et al., 2001. Forebrain and midbrain development requires epiblast-restricted Otx2 translational control mediated by its 3′ UTR. Dev. Camb. Engl. 128, 2989–3000. Bragg, D.C., et al., 2017. Disease onset in X-linked dystonia-parkinsonism correlates with expansion of a hexameric repeat within an SVA retrotransposon in TAF1. Proc. Natl. Acad. Sci. U. S. A. 114, E11020–E11028. Carri, A.D., et al., 2013. Developmentally coordinated extrinsic signals drive human pluripotent stem cell differentiation toward authentic DARPP-32+ medium-sized spiny neurons. Dev. Camb. Engl. 140, 301–312. Chambers, S.M., et al., 2009. Highly efficient neural conversion of human ES and iPS cells by dual inhibition of SMAD signaling. Nat. Biotechnol. 27, 275–280. Chen, C.-M., 2011. Mitochondrial dysfunction, metabolic deficits, and increased oxidative stress in Huntington's disease. Chang Gung Med. J. 34, 135–152. Costa, V., Scorrano, L., 2012. Shaping the role of mitochondria in the pathogenesis of

Acknowledgements We wish to thank the patients and probands for donating skin biopsies and Britta Meier as well as Andreas Niesel for their excellent technical support.

57

Experimental Neurology 308 (2018) 47–58

P. Capetian et al.

Nolte, D., Niemann, S., Müller, U., 2003. Specific sequence changes in multiple transcript system DYT3 are associated with X-linked dystonia parkinsonism. Proc. Natl. Acad. Sci. U. S. A. 100, 10347–10352. Precious, S.V., et al., 2016. FoxP1 marks medium spiny neurons from precursors to maturity and is required for their differentiation. Exp. Neurol. 282, 9–18. Rakovic, A., et al., 2018. Genome editing in induced pluripotent stem cells rescues TAF1 levels in X-linked dystoniaparkinsonism. Mov. Disord. Rembach, A., et al., 2004. Antisense peptide nucleic acid targeting GluR3 delays disease onset and progression in the SOD1 G93A mouse model of familial ALS. J. Neurosci. Res. 77, 573–582. Rocher, A.B., et al., 2016. Synaptic scaling up in medium spiny neurons of aged BACHD mice: a slow-progression model of Huntington's disease. Neurobiol. Dis. 86, 131–139. Rosales, R.L., 2010. X-linked dystonia parkinsonism: clinical phenotype, genetics and therapeutics. J. Mov. Disord. 3, 32–38. Sako, W., et al., 2011. Identification and localization of neuron-specific isoform of TAF1 in rat brain: implications for neuropathology of DYT3 dystonia. Neuroscience 189, 100–107. Segal, M., Greenberger, V., Korkotian, E., 2003. Formation of dendritic spines in cultured striatal neurons depends on excitatory afferent activity. Eur. J. Neurosci. 17, 2573–2585. Smith, G.A., et al., 2014. Progressive axonal transport and synaptic protein changes correlate with behavioral and neuropathological abnormalities in the heterozygous Q175 KI mouse model of Huntington's disease. Hum. Mol. Genet. 23, 4510–4527. Stanslowsky, N., et al., 2014. Functional differentiation of midbrain neurons from human cord blood-derived induced pluripotent stem cells. Stem Cell Res Ther 5, 35. Stanslowsky, N., et al., 2016a. Neuronal dysfunction in iPSC-derived medium spiny neurons from chorea-Acanthocytosis patients is reversed by Src kinase inhibition and F-actin stabilization. J. Neurosci. 36, 12027–12043. Stanslowsky, N., et al., 2016b. Functional effects of cannabinoids during dopaminergic specification of human neural precursors derived from induced pluripotent stem cells. Addict. Biol. http://dx.doi.org/10.1111/adb.12394. Takahashi, K., et al., 2007. Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell 131, 861–872. Toresson, H., Parmar, M., Campbell, K., 2000a. Expression of Meis and Pbx genes and their protein products in the developing telencephalon: implications for regional differentiation. Mech. Dev. 94, 183–187. Toresson, H., Potter, S.S., Campbell, K., 2000b. Genetic control of dorsal-ventral identity in the telencephalon: opposing roles for Pax6 and Gsh2. Dev. Camb. Engl. 127, 4361–4371. Tronnier, V.M., et al., 2015. Biochemical mechanisms of pallidal deep brain stimulation in X-linked dystonia parkinsonism. Parkinsonism Relat. Disord. http://dx.doi.org/10. 1016/j.parkreldis.2015.06.010. Ungrin, M.D., Joshi, C., Nica, A., Bauwens, C., Zandstra, P.W., 2008. Reproducible, ultra high-throughput formation of multicellular organization from single cell suspensionderived human embryonic stem cell aggregates. PLoS ONE 3, e1565. Vaine, C.A., et al., 2016. X-linked dystonia-parkinsonism patient cells exhibit altered signaling via nuclear factor-kappa B. Neurobiol. Dis. http://dx.doi.org/10.1016/j. nbd.2016.12.016. Vonsattel, J.P., et al., 1985. Neuropathological classification of Huntington's disease. J. Neuropathol. Exp. Neurol. 44, 559–577. Xu, C., et al., 2001. Feeder-free growth of undifferentiated human embryonic stem cells. Nat. Biotechnol. 19, 971–974. Zhou, S., et al., 2015. Neurosphere based differentiation of human iPSC improves astrocyte differentiation. Stem Cells Int. 2016 (e4937689). Zündorf, G., Reiser, G., 2011. Calcium dysregulation and homeostasis of neural calcium in the molecular mechanisms of neurodegenerative diseases provide multiple targets for neuroprotection. Antioxid. Redox Signal. 14, 1275–1288.

Huntington's disease. EMBO J. 31, 1853–1864. Domingo, A., et al., 2015. New insights into the genetics of X-linked dystonia-parkinsonism (XDP, DYT3). Eur. J. Hum. Genet. EJHG 23, 1334–1340. Domingo, A., et al., 2016. Evidence of TAF1 dysfunction in peripheral models of X-linked dystonia-parkinsonism. Cell. Mol. Life Sci. CMLS 73, 3205–3215. Durieux, P.F., Schiffmann, S.N., de Kerchove D'Exaerde, A., 2011. Targeting neuronal populations of the striatum. Front. Neuroanat. 5. Fan, M.M.Y., Raymond, L.A., 2007. N-methyl-D-aspartate (NMDA) receptor function and excitotoxicity in Huntington's disease. Prog. Neurobiol. 81, 272–293. Gécz, J., et al., 1999. Characterization of the human glutamate receptor subunit 3 gene (GRIA3), a candidate for bipolar disorder and nonspecific X-linked mental retardation. Genomics 62, 356–368. Gerfen, C.R., 1992. The neostriatal mosaic: multiple levels of compartmental organization. Trends Neurosci. 15, 133–139. Gittis, A.H., Kreitzer, A.C., 2012. Striatal microcircuitry and movement disorders. Trends Neurosci. 35, 557–564. Goldberg, J.A., Teagarden, M.A., Foehring, R.C., Wilson, C.J., 2009. Nonequilibrium calcium dynamics regulate the autonomous firing pattern of rat striatal cholinergic interneurons. J. Neurosci. 29, 8396–8407. Goto, S., et al., 2005. Functional anatomy of the basal ganglia in X-linked recessive dystonia-parkinsonism. Ann. Neurol. 58, 7–17. Goto, S., et al., 2013. Defects in the striatal neuropeptide Y system in X-linked dystoniaparkinsonism. Brain J. Neurol. 136, 1555–1567. Graham, R.K., et al., 2009. Differential susceptibility to excitotoxic stress in YAC128 mouse models of Huntington disease between initiation and progression of disease. J. Neurosci. 29, 2193–2204. Herzfeld, T., et al., 2013. X-linked dystonia parkinsonism syndrome (XDP, lubag): diseasespecific sequence change DSC3 in TAF1/DYT3 affects genes in vesicular transport and dopamine metabolism. Hum. Mol. Genet. 22, 941–951. Hollmann, M., Hartley, M., Heinemann, S., 1991. Ca2+ permeability of KAAMPA—gated glutamate receptor channels depends on subunit composition. Science 252, 851–853. Hussein, S.M., et al., 2011. Copy number variation and selection during reprogramming to pluripotency. Nature 471, 58–62. Imamura, K., et al., 2016. Calcium dysregulation contributes to neurodegeneration in FTLD patient iPSC-derived neurons. Sci. Rep. 6. Ito, N., et al., 2016. Decreased N-TAF1 expression in X-linked dystonia-parkinsonism patient-specific neural stem cells. Dis. Model. Mech. 9, 451–462. Jayakar, S.S., Dikshit, M., 2004. AMPA receptor regulation mechanisms: future target for safer neuroprotective drugs. Int. J. Neurosci. 114, 695–734. Kaufman, A.M., et al., 2012. Opposing roles of synaptic and extrasynaptic NMDA receptor signaling in cocultured striatal and cortical neurons. J. Neurosci. 32, 3992–4003. Kelsom, C., Lu, W., 2013. Development and specification of GABAergic cortical interneurons. Cell Biosci. 3, 19. Kuijlaars, J., et al., 2016. Sustained synchronized neuronal network activity in a human astrocyte co-culture system. Sci. Rep. 6. Lee, L.V., et al., 2011. The unique phenomenology of sex-linked dystonia parkinsonism (XDP, DYT3, ‘Lubag’). Int. J. Neurosci. 121 (Suppl. 1), 3–11. Li, X.-J., et al., 2009. Coordination of sonic hedgehog and Wnt signaling determines ventral and dorsal telencephalic neuron types from human embryonic stem cells. Dev. Camb. Engl. 136, 4055–4063. Makino, S., et al., 2007. Reduced neuron-specific expression of the TAF1 gene is associated with X-linked dystonia-parkinsonism. Am. J. Hum. Genet. 80, 393–406. Martynoga, B., Morrison, H., Price, D.J., Mason, J.O., 2005. Foxg1 is required for specification of ventral telencephalon and region-specific regulation of dorsal telencephalic precursor proliferation and apoptosis. Dev. Biol. 283, 113–127. Marx, U., et al., 2013. Automatic production of induced pluripotent stem cells. Procedia CIRP 5, 2–6.

58