Higher Vulnerability and Stress Sensitivity of Neuronal Precursor Cells Carrying an Alpha-Synuclein Gene Triplication

Parkinson disease (PD) is a multi-factorial neurodegenerative disorder with loss of dopaminergic neurons in the substantia nigra and characteristic intracellular inclusions, called Lewy bodies. Genetic predisposition, such as point mutations and copy number variants of the SNCA gene locus can cause very similar PD-like neurodegeneration. The impact of altered α-synuclein protein expression on integrity and developmental potential of neuronal stem cells is largely unexplored, but may have wide ranging implications for PD manifestation and disease progression. Here, we investigated if induced pluripotent stem cell-derived neuronal precursor cells (NPCs) from a patient with Parkinson’s disease carrying a genomic triplication of the SNCA gene (SNCA-Tri). Our goal was to determine if these cells these neuronal precursor cells already display pathological changes and impaired cellular function that would likely predispose them when differentiated to neurodegeneration. To achieve this aim, we assessed viability and cellular physiology in human SNCA-Tri NPCs both under normal and environmentally stressed conditions to model in vitro gene-environment interactions which may play a role in the initiation and progression of PD. Human SNCA-Tri NPCs displayed overall normal cellular and mitochondrial morphology, but showed substantial changes in growth, viability, cellular energy metabolism and stress resistance especially when challenged by starvation or toxicant challenge. Knockdown of α-synuclein in the SNCA-Tri NPCs by stably expressed short hairpin RNA (shRNA) resulted in reversal of the observed phenotypic changes. These data show for the first time that genetic alterations such as the SNCA gene triplication set the stage for decreased developmental fitness, accelerated aging, and increased neuronal cell loss. The observation of this “stem cell pathology” could have a great impact on both quality and quantity of neuronal networks and could provide a powerful new tool for development of neuroprotective strategies for PD.


Introduction
Two of the most critical parameters determining cellular functionality and health are energy generation and distribution, because they are the driving force behind all biological processes [1]. Mitochondria are at the center of cellular metabolism and energy-dependent signaling processes [2], compartmentalizing cellular bioenergetic pathways and linking cellular energy to gene expression [3]. Thus, mitochondria are directly involved in regulation of cell fate and neuroplasticity.
In the process of neuronal degeneration, mitochondria are central regulators of cellular fate and apoptosis [4]. It is not surprising then that impaired mitochondrial function has been shown in several neurodegenerative disorders, including Parkin-son's disease (PD). Interestingly, mitochondrial dysfunction has been implicated in various inherited forms of parkinsonism as well (e.g. Parkin, PINK1, LRRK2, and SNCA genes) [5]. Apart from gene defects, environmental toxicants directly affecting mitochondrial function have been identified as risk factors for PD etiology [6].
The a-synuclein (SNCA) gene was the first gene linked to familial PD [7], and genetic alterations at the SNCA locus are one of the most significant genetic risk factors for development of sporadic forms of the disease [4,8,9]. The SNCA triplication (SNCA-Tri), resulting in an overexpression of wildtype a-synuclein protein (a-syn) which leads to early onset progressive parkinson-ism, represents an ideal system to investigate synucleinopathyspecific disease mechanisms [10,11].
a-syn is associated with mitochondria [12], and pathological asyn oligomers at mitochondrial membranes exhibit direct mitochondrial toxicity [13,14], affect mitochondrial dynamics [15,16] and interaction with other organelles [17]. In skin fibroblasts from the patient described in this study, we found significant changes in ATP production, reduction in mitochondrial membrane potential and complex I activity. The SNCA-tri fibroblasts were also more sensitive to oxidative stressors. The phenotype could be partially reversed by siRNA knockdown of a-syn which suggests a direct causative role for increase concentrations of intracellular of a-syn [18].
Mitochondrial integrity and functional metabolism are also essential for stem cell proliferation [19,20] as well as differentiation [21][22][23]. Until now, cell biology and metabolic properties of human neural stem cells carrying PD mutations have not been thoroughly investigated, and even less is known about the impact of environmental factors on these cells.
In this study we take advantage of this new technology to investigate the impact of excess intracellular concentrations of asyn (using tissue from our patient with a SNCA gene triplication (SNCA-Tri)) on cellular and mitochondrial function of human iPSC-derived neuronal precursor cells (NPCs). Using semiquantitative and high throughput screening (HTS) and high content imaging (HCI) technologies in addition to biochemical assays and conventional imaging, we observed that there are profound effects on mitochondria function and energy production in this novel model of PD. As iPSC-derived NPC populations can give rise to a variety of neuronal cell types, our studies may warrant further investigations on how decreased developmental fitness, accelerated aging and increased susceptibility to stress contribute to neurodegenerative processes that set the stage for to neuronal cell loss in PD.

Generation of patient-derived induced pluripotent stem cells (iPSC)
We generated multiple iPSC clones from fibroblasts from a patient with parkinsonism carrying a SNCA triplication and a four year older mutation-negative healthy female sibling [24]. Two iPSC clones with normal karyotype were used for further analyses. As an additional control, an iPSC line from a healthy age and sexmatched mutation-negative individual was generated. The study (protocol ECH-08-20) was approved by a local ethics committee (El Camino Hospital, Mountain View, CA), reviewed annually, and all participants signed informed consent.
For experiments with toxicants and inhibitors, NPCs were cultured under the following conditions for 18 hrs unless specified differently in the experimental procedures: Standard NPC growth medium as above (HG = high glucose), HG plus 20 mM rotenone (Sigma #R8875) (HG+R) or modified NPC growth medium without glucose (NG) (Life Techn. Neurobasal-A, Formula 05-0128DJ). For experiments determining Reactive oxygen species (ROS) generation, B27 was replaced with the formulation without antioxidants (Life Tech, #1889-038).

Cell viability and survival
NPCs on Geltrex-coated 96-well plates were challenged as described under method section: NPC toxicant and inhibitor treatment, with the exception that 20 mM paraquat was applied. Medium was changed daily and cells were imaged with Zeiss Axiovert 25 microscope (with CP-Achromat 10x/0.25 Ph1 objective) and Canon EOS400 camera. Images were contrast enhanced, converted to grayscale and number of cells per image frame was analyzed with ImageJ [36].

Plate reader HTS
For HTS plate reader analyses, several independent experiments (as specified in results) with 3-4 replicates per cell line and treatment regimen were conducted. Signals from 96-well plates (integration time 1 sec) were acquired by bottom read in orbital mode (8-spot measurements with 2 mm radius). Kinetic measurements were recorded over 30 min with a sampling rate of 1 read/ min. Data analysis was performed with BMG Mars software.

HCI microscopy
HCI microscopy was performed with an ImageXpress Micro System, Molecular Devices LLC and HTS was performed with a multi-wavelength plate reader (Polarstar Omega, BMG) or by flow cytometry (Accuri C6 with C-sampler, BD Biosciences). Images from NPCs seeded in 96-well plates were acquired with 10x and 20x Plan Fluor objectives (4 replicates with 4 sites per well) using Ex./Em. 346/480 nm, 490/530 nm and 535/585 nm filter sets and then analyzed with the MetaXpress software and the following modules: Cell Scoring (cell metrics), Granularity (endpoint analysis of organelle and mitochondrial metrics) and Transfluor (colocalization of two or more fluorophores).

Mitochondrial superoxide
Adherent NPCs in 96-well plates were incubated with 2 mM MitoSOX (Ex./Em. 510/580 nm) (Life Techn. #M36008) in HBSS plus for 10 min at 37uC in the dark. Cells were then washed, MitoSOX fluorescence was detected at Ex./Em. 544/590610 nm and normalized to cell number as under method section: ROS.

Caspase activation
Caspase activity was assayed from NPCs seeded in 96-well plates with the EnzChek Caspase-3 Assay Kit (Life Techn. #E-13183). Apoptosis was induced with 4 mM SP for 4 hrs. After cell lysis, an aliquot was removed for protein (BCA-Assay) and 20 mM Z-DEVD-AMC substrate (Ex./Em. 342/441 nm) was added. After 30 min incubation at RT, fluorescence of the cleaved Z-DEVD-AMC substrate was acquired (Ex./Em. 380/445 nm) and normalized to AMC standard and protein content. Change in caspase activity (D mM AMC/mg protein) in NPCs was measured by detaching cells from substrate and permeabilization with 50 mg/ml digitonin (Sigma, #D5628) in HBSS plus for 10 min at RT. After addition of caspase substrate, fluorescence was recorded every 30 min for 4 hrs.

Metabolic flux analysis
Cellular response to specific mitochondrial inhibitors was analyzed on a Seahorse XF24 by monitoring O 2 consumption (OCR) and extracellular acidification rates (ECAR). Three days before the assay 1610 5 cells were seeded on V7 24-well plates (Seahorse Bioscience, #SEA100777004) pre-coated with 150 mg/ ml Geltrex. Culture medium was changed to Seahorse Assay Medium (#SEA102353100) supplemented with 10 mM pyruvate and 25 mM glucose one hour before starting the measurements. The neurotoxin 6-hydroxydopamine (6-OHDA) at 250 mM final concentration (f.c.) was added to half of the wells. OCR and ECAR baselines were recorded three times before the consecutive addition of (all f.c.): 1 mM oligomycin, 1.5 mM CCCP and 5 mM rotenone (Rot) also containing 1 mM antimycin A (Ant). Control wells without addition of drugs were recorded in every plate as a viability reference for each cell line and used to correct the values of spare respiratory capacity (OCR CCCP -OCR basal ) and proton leakage (OCR Oligomycin -OCR Rot+Ant ).
For determination of metabolic activity, membrane asymmetry and membrane permeability, NPCs were assayed with C 12 -Resazurin [37]  ATP production 2610 4 cells/mm 2 cells seeded in 96-well plates 48 hrs before the experiment were treated as described under ''NPC Treatment'', then lysed on ice and assayed with a coupled luciferin/luciferase assay (ATPLite, Perkin Elmer, #6016941) according to the manufacturer's instructions. Plates were read on a VERITAS Luminometer (Turner Biosystems #998-9100). ATP content as determined by an ATP standard was then normalized to protein content determined by BCA Assay.

Proteasome Activity
Protein aggregation in NPCs was assayed in adherent NPCs cultured with 20 mM rotenone alone or with 10 mM of the proteasome inhibitor MG132 (ProteoStat aggresome detection Kit, Enzo Life Sciences #ENZ-51035-K100). NPCs, either on slides (for fluorescence microscopy) or detached from substrate (for flow cytometry) were fixed and permeabilized and then stained with 5 mM of the aggresome/proteasome specific dye Bodipy TMR-AHX3L3VS (Ex./Em. 500/600 nm) and Hoechst 33342 according to the manufacturer's instructions. From flow cytometry of NPCs, aggregesome propensity factors (APF) were calculated from the mean RFU (MRFU) of Bodipy-TMR fluorescence: APF = 1006[MRFU MG132 treated cells2MRFU untreated cells]/MRFU MG132 treated cells.

Cytochrome c release
Subcellular fractionation of cultured NPCs was done as previously described [40]. All steps were performed at 4uC. 1610 7 cells treated with 4 mM PQ for 1 hr were harvested and resuspended in 500 ml of mitochondrial isolation buffer (250 mM sucrose, 1 mM EDTA, 1 mM DTT, 10 mM HEPES, pH 7.5) supplemented with protease inhibitor (Complete Ultra, Roche #11836153001) and 1 mM DNase I (Roche #03724778103). Cells were fractionated by 3 passages through a 30G needle, centrifuged at 1000 g for 10 min. The supernatant was then spun at 10.000 g for 15 min to collect cell organelles and the supernatant was concentrated with a 3 kD Microcon filtration unit (YM-3 Millipore #42420) at 14.000 g for 30 min. Protein concentration was determined by BCA-Assay and 30 mg protein were used for immunoblotting.

Statistical analysis
For individual experiments the mean 6 SD or mean 6 SEM were plotted. Non-directional Student's t test was used for direct statistical comparisons. For multiple comparisons, multivariant ANOVA was used. Where significant F-values were obtained, pair wise comparisons were made using Wilcoxon-Mann/Whitney post hoc analysis. Differences were considered statistically significant at p#0.05.

Results
Normal morphology but reduced proliferative capacity in SNCA-Tri NPCs Patient-derived NPC lines carrying the SNCA gene triplication (SNCA-Tri), the SNCA-Tri shRNA knockdown line (SNCA-Tri KD) and those from two unaffected controls (Ctrl) were morphologically indistinguishable under normal, high glucose growth conditions (HG) ( Figure 1A) and did not show differences in mitochondrial shape or content ( Figure 1B). All three NPC lines expressed neuronal stem cell markers Nestin and SOX1 ( Figure 1C). Confluent cultures of SNCA-Tri NPCs expressed twice as much a-syn compared to controls irrespective of passage number, and SNCA shRNA knockdown resulted in a significant reduction of SNCA expression levels in NPCs (n = 4, mean 6 SEM, SNCA-Tri/Ctrl: 12.4/5.9, SNCA-Tri KD: 8.3, ***p# 0.001, t-test) ( Figure 1D). The differential a-syn expression levels in NPCs were also observed by ICC, with the protein associated with sub-cellular structures such as with mitochondria Changes in cell cycle have been observed in other progenitor cell types used in models for neurodegeneration [41]. Under naïve conditions, cell cycle analysis showed a reduction in S-Phase DNA content of SNCA-Tri NPCs (n = 3, mean 6 SD, 24.4%/15.7%,  (Figure 2A).
To investigate the impact of starvation and toxicants on SNCA-Tri NPC viability and stress resistance, we established protocols to imitate general metabolic and specifically mitochondrial stress conditions. We developed a starvation protocol that omitted glucose from the culture medium and a toxicant stressor panel that exposed NPCs to environmental toxins and toxicants such as rotenone (R), paraquat (PQ), staurosporine (SP) and oligomycin (O).

Altered cellular energy balance and decreased mitochondrial function in SNCA-Tri NPCs
We next investigated energy status and metabolism in SNCA-Tri NPCs. When SNCA-Tri NPCs challenged with rotenone and loaded with tetramethylrhodamine (TMRM) were examined by fluorescence microscopy, they showed reduced mitochondrial membrane potential (MMP) ( Figure 3A). We then quantified MMP by HTS fluorescence plate reader analysis using the ratiometric fluorescent dye JC-10. MMP in SNCA-Tri NPCs was significantly more impaired by rotenone and glucose withdrawal compared to controls, and knockdown of a-syn reinstated normal MMP levels in SNCA-Tri KD NPCS (n = 8, mean 6 SEM, Ctrl/ SNCA-Tri/SNCA-Tri KD for HG+R: 202/29/194 (x10 4 ), *p# 0.05; for NG: 92/30/118 (x10 3 ) **p#0.006) ( Figure 3B). In addition, when monitoring MMP over time, SNCA-Tri NPCs lost their MMP significantly faster than Ctrl NPCs when exposed to stress by extended HTS analysis (1 hr), which was compounded by rotenone stress and glucose withdrawal and could be returned to rates in Ctrl by knockdown of a-syn  Figure 3D).
To investigate mitochondrial function in live NPCs we measured cell respiratory control in live NPCs ( Figure 3E). SNCA-Tri NPCs showed altered O 2 consumption rates and nonmitochondrial respiration (Ctrl/SNCA-Tri: 24%/8%), and OCR was sensitive to the neurotoxin 6-hydroxy-dopamine (6-OHDA). Relevant for mitochondrial function, spare respiratory capacity, representing the ability of mitochondria to respond to an increase in energy demand (expressed as the quantitative difference between basal OCR and maximal uncontrolled OCR) by addition of the uncoupler CCCP was reduced from 57% in control NPCs to 34% in SNCA-Tri NPCs (n$17, mean 6 SEM, (#p,0.05). When NPCs were exposed to the neurotoxin 6-OHDA their spare respiratory capacity was further decreased, with the difference between Ctrl and SNCA-Tri NPCs being preserved (37%/14%). Mitochondrial proton leak, represented by the reduction of OCR in presence of oligomycin and indicative of mitochondrial uncoupling from respiration, was significantly higher in SNCA-Tri NPCs (OCR Ctrl/SNCA-Tri: 48%/32%), supporting the observations of decreased coupling efficiency and less stable MMP in SNCA-Tri NPCs.
The glycolytic activity of both lines (measured by medium acidification and graphed as extracellular acidification rate (ECAR)) ( Figure 3F) was similar in Ctrl and SNCA-Tri NPCs, as indicated by comparable levels of lactic acid formation after inhibition of mitochondrial ATP production, but ECARs in SNCA-Tri cells stressed by 6-OHDA were significantly decreased, confirming the findings from the plate reader based HTS screen for ATP levels.

Delayed protein import and increased protein aggregation in SNCA-Tri NPCs
Protein biosynthesis and import into cellular organelles are essential for cellular anabolic metabolism [42]. When we transduced adherent NPCs with viral BacMam vectors encoding mitochondrial-and peroxisomal-targeted fluorescent proteins, we observed both delayed appearance and reduced levels of organelle-specific fluorescence in SNCA-Tri NPCs ( Figure 4A/ B).
On the catabolic side, dysfunction of the mechanisms to repair and remove abnormal proteins, such as impaired unfolded protein response and proteasome function, have been shown to play a pivotal role in PD disease progression [43,44]. When we analyzed cytoplasmic protein aggregates (aggresomes) in proteasome-inhibitor treated SNCA-Tri and Ctrl NPCs by microscopy and semiquantitative flow cytometry ( Figure 4C, D), we observed a significant increase in aggresome formation in rotenone and proteasome inhibitor-treated SNCA-Tri cells ( = 3, mean 6 SD, APF Ctrl/SNCA-Tri: 51/120, *p = 0.041), suggesting increased activity of the cellular proteasome system.
Increased cellular superoxide production has also been implicated in PD pathology [46]. HTS analysis of NPCs labeled with a mitochondria-specific fluorescent superoxide indicator showed increased superoxide levels in SNCA-Tri NPCs that was compounded by exposure to cellular stress and was reduced by knockdown of a-syn (n = 4, mean 6 SD, Ctrl/SNCA-Tri/SNCA-  Figure 5D). The higher mitochondrial superoxide levels were also confirmed by flow cytometry analysis using the same exposure conditions (data not shown). These results suggest defects in scavenging and clearing cellular ROS and/or mitochondrial superoxide levels in SNCA-Tri NPCs, or may indicate a more complex regulation and balance of ROS and superoxide levels in neural progenitor cells [47].

Impaired mitochondrial integrity and permeability transition in SNCA-Tri NPCs
Opening of the mitochondrial permeability transition pore (MPT) results in mitochondrial depolarization, uncoupling of oxidative phosphorylation, large-amplitude mitochondrial swelling and ultimately apoptosis [23]. We monitored mitochondrial integrity and MPT opening in live NPCs by loading mitochondria with fluorescent calcein in presence of a cytoplasmic calcein quencher and conducted both endpoint analysis of mitochondrial calcein levels and kinetic studies on the rate of mitochondrial calcein loss. HCI analysis revealed higher mitochondrial calcein loading capacity in SNCA-Tri NPCs challenged with rotenone (n = 8, mean 6 SD, Ctrl/SNCA-Tri: 3.4/4.9, *p = 0.039), indicative of increased cellular and mitochondrial stress (Figure 6A). To investigate the mitochondrial resilience to toxininduced MPT opening, we analyzed mitochondrial calcein fluorescence in NPCs after treatment with 4 mM staurosporine by HCI. The SNCA-Tri NPC populations showed a significantly reduced mitochondrial calcein signal, indicative of more pronounced MPT opening compared to Ctrl NPCs (n = 3, mean 6 SD, Ctrl/SNCA-Tri, HG: 834/457, HG+R: 1425/1011, NG: 864/574, HG+Iono: 187/190, *p#0.01) under all experimental conditions ( Figure 6B, C).
HCI analysis of the rate of mitochondrial calcein loss revealed a significantly faster loss of calcein signal in SNCA-Tri cells regardless of treatment regimen (n = 8, mean 6 SD, Ctrl/ SNCA-Tri, HG: 20.06/20.12, HG+R: 20.17/20.28, HG+ Iono: 20.03/20.04, *p#0.01) that was concurrent with reduction in TMRM signal (data not shown) ( Figure 6D). Taken together, these observations indicate both increased MPT opening and a faster rate of mitochondrial permeabilization in SNCA-Tri NPCs.
To confirm the loss of mitochondrial outer membrane integrity in toxicant exposed SNCA-Tri cells, we challenged NPCs with PQ to induce acute cell death and then performed cytochrome c immunocytochemistry. SNCA-Tri cells on average showed significantly reduced cytochrome c signal, confirming outer mitochondrial membrane permeabilization in these cells (Figure 6E). The increased mitochondrial cytochrome c loss in SNCA-Tri NPCs under these conditions was also confirmed by immunoblot analysis of cytochrome c content in cytosol-and organelle-enriched sub-cellular fractions from PQ treated NPCs. ( Figure 6F).

Caspase activation and apoptosis
As we had observed significant deficiencies in cellular catabolic and anabolic processes, energy metabolism and altered ROS levels as well as increased membrane asymmetry and cell death in challenged SNCA-Tri NPCs (see Figure 2D), we investigated the mechanism of NPC cell death.
To better resolve temporal caspase activation in NPCs, we examined time resolved caspase 3 cleavage activity in permeabilized NPCs after 15 min exposure to staurosporine ( Figure 7B). Under normal growth conditions (HG), both NPC lines showed stable caspase 3 activation during the first 60 min., with the SNCA-tri NPCs displaying higher levels of activation during this time. Under rotenone stress, both cell lines, but particular the SNCA-Tri NPCs displayed high caspase activation levels that decreased more rapidly, suggesting accelerated loss of cellular function compared to control and confirming the observations about faster loss of viability in SNCA-Tri NPCs.

Discussion
The insight into molecular pathomechanisms responsible for neurodegeneration of Mendelian forms of parkinsonism is advancing rapidly, however little is known about the impact of these genetic defects on function and fate of human stem or neuronal precursor cells and their predisposition to nutritional and environmental stress. A few reports showed an impairment of embryonic neurogenesis in animal models for PD [48,49]. and in a comprehensive screen in iPSC-derived neurons carrying mutations in PD genes has shown convergence of cellular disease mechanisms, such as increased cellular stress and mitochondrial dysfunction in these neurons [25]. However, the impact of these mutations on function, viability, and proliferative potential of human NPCs has not yet been investigated.

SNCA-Tri NPCs show reduced viability and metabolic capacity
To model the impact of nutritional and environmental challenges on the cellular networks and particularly mitochondria, we subjected NPCs to metabolic stress by exposure to the mitochondrial toxicant rotenone, free radical generators such as paraquat, or nutrient (glucose) deprivation. We observed slower proliferative capacity, reduced viability and decreased cell survival rates in SNCA-Tri NPCs under cellular stress conditions, which was alleviated by SNCA shRNA knock down in these NPCs, suggesting a direct impact of a-syn overexpression on cell cycle progression, cell survival, metabolic fitness and stress resistance.
Cellular stress significantly affected cellular ATP content and mitochondrial membrane potential (MMP) in SNCA-Tri NPCs, suggesting a reduced capacity of these NPCs to mitigate metabolic challenges and maintain mitochondrial functionality.
Metabolism in highly proliferative cells such as NPCs is tightly regulated, with strong control of cytoplasmic glycolysis and mitochondrial OXPHOS [50]. The observed changes in cellular ATP content in SNCA-Tri NPCs together with mitochondrial  Metabolic flux analysis, measuring oxygen consumption rate and extracellular acidification, confirmed the mitochondrial membrane instability. The observed media acidification suggest changes in glycolytic flux in these NPCs, possibly leading to an imbalance of pyruvate and lactate and resulting in metabolic acidosis [51].
Taken together, these results point to de-synchronization of glycolytic and mitochondrial energy metabolism, resulting in reduced viability and proliferative capacity, supporting our previous findings in skin fibroblast cultures of this patient .
Respiratory chain function has consistently been found impaired in PD and in association with a-syn pathophysiology [15], and the observed association of a-syn with subcellular lipid bilayers suggests a general effect of overexpression on mitochondrial protein import and possibly respiratory chain functionality. This hypothesis is supported by the delayed appearance of transgenic fluorescent protein in SNCA-Tri NPC mitochondria, suggestive of altered mitochondrial protein import and assembly that could be a contributing factor to mitochondrial respiratory chain dysfunction.

SNCA-Tri NPCs display increased cellular stress and ROS susceptibility
The ubiquitin proteasome system (UPS) is recycling dysfunctional cellular proteins [52], and impairment of this system plays an important role in neurodegenerative processes associated with PD [43]. Increased aggresome formation in toxicant-treated SNCA-Tri NPCs suggests that the UPS and proteasome activity may be heavily used, thus taxing the cellular catabolic systems. These data are also supported by the strong activity of peroxisomal matrix protein import observed in these NPCs.
Disturbed mitochondrial respiratory chain assembly and function result in increased ROS and superoxide production [53]. Similarly, altered protein transport and turnover result in increased oxidative stress in iPS-derived neuronal cultures [24,26]. Increased ROS levels are consistently reported in PD pathophysiology [13], and we observed both higher steady state ROS levels and increased ROS production in SNCA-Tri NPCs; under normal growth conditions, under toxicant exposure, and nutrient deprivation. Our findings point towards increased ROS formation as consequence of the physiological changes observed in SNCA-Tri NPCs. In addition to the above direct ROS effects on mitochondrial function, the increased a-syn levels in SNCA-Tri NPCs could also impair intracellular ROS regulation by affecting mitochondrial anti-oxidant defense mechanisms, cellular RedOx balance and cell signaling [54].
Our data support a model where a-syn overexpression-initiated changes impair regulation and efficiency of cellular energy generation. These changes then lead to increased cellular stress, increased cellular ROS production and altered RedOx balance that could also affect NPC growth and differentiation.
analysis. NPCs treated and prepared as under B) were loaded with 4 mM stauropsporine and changes in calcein signal normalized to cell number and mitochondrial content (D Norm. RFU) were recorded every 1 min for 20 min (n = 8, mean 6 SD, Ctrl/SNCA-Tri, HG: 20.06/20.12, HG+R: 20.17/20.28, HG+Iono: 20.03/20.04, *p#0.01). E) Cytochrome c immuno-cytochemistry in Ctrl and SNCA-tri NPCs challenged with 200 mM paraquat (PQ) 15 min. before fixation. Shown are permeabilized cells probed with cytochrome c antibody, detected by an Alexa-488 nm labeled secondary antibody (green). Cells were counter stained with Hoechst 33342 (blue) (Scale bar: 100 mm, insert: 10 mm). F) Immunoblot analysis of cytochrome c levels in sub-cellular fractions containing either cellular organelles (containing bound cytochrome c) or cytosolic proteins (with soluble cytochrome c) from NPC cell lysates (Ctrl and SNCA-Tri) treated with paraquat (PQ) as under E). Cytochrome c (14 kDa) and GAPDH (40 kDa) specific antibodies were detected by a secondary IR-dye conjugate. doi:10.1371/journal.pone.0112413.g006 Figure 7. Apoptosis sensitivity and caspase activation. A) Caspase 3 activity in cell lysates from adherent NPCs either left untreated or treated with 20 mM rotenone (R) for 18 hrs and then exposed to 1 uM staurosporine for 120 min before analysis. HTS analysis for caspase 3 activity from cell lysates was by activation of the fluorescent caspase substrate 7-amino-4-methylcoumarin (AMC) (Ex./Em. 340/440 nm) (n = 9, mean 6 SEM, Ctrl/SNCA-Tri/SNCA-Tri KD, HG: 33/69/42, HG+R: 42/129/87, NG: 55/138/85, *p#0.050, **p#0.0035; from three independent experiments). B) Kinetics of caspase 3/7 activity in permeabilized NPCs pretreated as described under B) and assayed 15 min after staurosporine treatment. Changes in caspase 3 activity are depicted as DmM AMC fluorescence/min + mg cellular protein (detected by Bradford protein assay) (n = 9, mean 6 SEM). doi:10.1371/journal.pone.0112413.g007 Role of Alpha-Synuclein in Neuronal Precursor Cells Increased mitochondrial apoptosis sensitivity of SNCA-Tri NPCs Cellular stress is ascribed to induce transient or irreversible MTP opening, with the latter leading to initiation of mitochondria-mediated apoptosis. Under acute cellular stress, SNCA-Tri mitochondria showed hyperpolarization and increased mitochondrial calcein loading (indicative of increased calcium levels), which resulted in more rapid MTP opening. Mitochondria in connection with the endoplasmic reticulum are important components of cellular Ca 2+ regulation [55], and dysregulation of this system has been shown to be part of the molecular pathomechanism in PD [56]. A-syn has been shown to increase mitochondrial Ca 2+ uptake and impairs mitochondrial function [17], ultimately resulting in Ca 2+ mediated mitochondrial apoptosis [57].
Mitochondrial toxicants such as rotenone raise mitochondrial ROS levels and increase mitochondrial Ca 2+ stress [58]. In cell signaling, Ca 2+ and Ca 2+ -dependent pathways regulate components of ROS/RedOx homeostasis [59], mitochondrial metabolic rate and ROS generation [60]. The higher cellular ROS burden, lower energy levels and higher propensity for MTP opening in connection with Ca 2+ stress demonstrate the increased apoptosis sensitivity in SNCA-Tri NPCs [2]. As stem cells have a decreased capacity to mediate larger Ca 2+ fluctuations and require tight control of ROS/RedOx during the onset of differentiation [61], abnormal Ca 2+ homeostasis may also impact proliferation and differentiation of SNCA-Tri stem cells [62].

Caspase Activation
Caspases have an important role both in the initiation of apoptotic events [63,64], but also in sensitizing and priming cells to developmental changes [65]. When cultured under starvation conditions, SNCA-Tri NPCs displayed greater loss of viability and cell membrane integrity in response to mitochondrial toxicant treatment. Investigation of caspase activation patterns in toxicant and ROS challenged NPCs, showed significantly higher caspase activation in SNCA-Tri NPCs with functionally impaired mitochondria. In this context, our results also demonstrate the mitochondrial control of apoptosis initiation in NPCs. In unaffected NPCs, general and mitochondrial stress can be compensated by activation of effective anti-apoptotic mechanisms, whereas in SNCA-Tri cells with impaired metabolism and stress resistance, these challenges result in an accelerated induction of cell death. Considered the role of caspases in development, the altered caspase expression patterns in challenged SNCA-Tri NPCs could also affect regulation of NPC differentiation.

Conclusion and Outlook
In ''a-synucleinopathic'' stem cells, a-synuclein overexpression affects cellular and metabolic plasticity, increases oxidative stress, lowers the reserve energy capacity and results in greater sensitivity for apoptosis and increased caspase activation under cellular stress. Our studies also suggest the existence of a stress threshold in NPCs, with insults and challenges (such as starvation and toxicant exposure) exceeding this threshold in SNCA-Tri NPCs at an lower concentrations and resulting in the manifestation of a ''stem cell pathology'' [66]. Our findings also raise new questions about the role of a-synuclein in regulation of mitochondrial activity in neuronal stem cells [67]. As the modulation of energy metabolism and metabolic signaling processes are essential for cell plasticity and cell fate decisions, our observations may have implications concerning the ability of PD-patient derived NPCs to form fully functional neuronal networks.