The application of iPSCs in Parkinson’s disease

The discovery and application of induced pluripotent stem cells (iPSCs) provide a novel treatment modality for diseases, which remain incurable. Particularly, in the treatment of neurodegenerative diseases such as Parkinson’s disease (PD), iPSC‐technology holds an interesting prospect for replacement therapy. Currently, the prognostic improvement of PD is limited and relies on symptomatic treatment. However, the symptomatic dopamine‐replacement therapies lose their long‐duration responses, and novel regenerative treatment modalities are needed. Animal models have provided valuable information and identified pathogenic mechanisms underlying PD but the lack of models that recapitulate the complex pathophysiology of the disease postpones further development of novel therapeutics. This review summarizes the possible uses of iPSCs in PD and discusses the future investigations needed for iPSCs as a possible treatment of PD patients.


INTRODUCTION
The medical research and application of induced pluripotent stem cells (iPSCs) is advancing rapidly. Although iPSCs are potentially applicable in several different fields of medical research and for clinical purposes (Takahashi and Yamanaka, 2006), they have proven to be particularly valuable in the research of neurodegenerative disease research e.g., Parkinson's disease (PD) (Zhu et al., 2019). PD is characterised by the degeneration of the midbrain dopaminergic (mDA) neurons that are essential for voluntary motor control. The replacement of the lost mDA neurons with iPSCs differentiated into mDA progenitors has shown promising potential in re-establishing DA function (Caiazzo et al., 2011;Pfisterer et al., 2011). Moreover, neural biopsies from PD patients are often inaccessible and difficult to obtain in sufficient quantities for in vitro studies of the underlying pathology (Gibb and Lees, 1988;Jowaed et al., 2010). Biopsies often represent later dis-ease stages of PD, which complicates the recapitulation of the comprehensive neuropathology of the disease. However, with iPSC-technology, it is possible to generate and culture patient-specific neurons in vitro from biopsy specimens and obtain a model system for investigation of e.g., early stages of PD (Devine et al., 2011). Furthermore, the technology provides possibilities for novel drug screening and it advances the field of personalized medicine (Plummer et al., 2019). Since the discovery of iPSCs, which was made by Nobel laureate Professor Shinya Yamanaka (Takahashi and Yamanaka, 2006;Takahashi et al., 2007) more than a decade ago, the technology has advanced significantly. Professor Yamanaka's research has undoubtedly contributed to all fields of science; however, primarily three applications are exceptionally relevant for iPSC-technology. These comprise novel drug development, regenerative therapy and disease modelling. Although the discovery of iPSC-technology has transformed biomedical research and created novel possibilities for cell identi-ty manipulation (Scudellari, 2016), challenges are still ahead. This review will address and discuss the current applications of iPSCs and the limitations of iPSCs as a restorative therapy for PD.

iPSC-technology
iPSCs constitute of a group of reprogrammable pluripotent stem cells (PSCs), which are derived from adult cells. PSCs are undifferentiated and have the unique capacity to divide endlessly and thereby self-renew. Furthermore, PSCs can develop into the three primary germ cell layers of the early embryo (endoderm, ectoderm and mesoderm) and hereby differentiate into all 230 different cell types of the adult human body (Morgani et al., 2018).
In 2006, an adult mouse cell, a fibroblast, was converted into a PSC (Takahashi and Yamanaka, 2006). A year later, the same method was proven possible in humans as well (Takahashi et al., 2007;Yu et al., 2007). The somatic cells were reprogrammed using retroviral vectors to express specific transcription factors, which are active in embryonic stem cells (ESCs) but are suppressed after cell differentiation. The applied transcription factors (Yamanaka factors) were c-Myc, Klf4, Oct3/4 and Sox2 (Takahashi and Yamanaka, 2006). Oct4 and Sox2 are still used to this day, while improved reprogramming circumvented the need for the oncogene c-myc (Miyoshi et al., 2011;Habib et al., 2013). The differentiation process is well monitored and the technology must follow the Guidelines for Stem Cell Research and Clinical Translation (Daley et al., 2016). The survival of the iPSCs depends on various growth factors such as fibroblast growth factor 2 (FGF-2) and the iPSCs are kept in an undifferentiated state due to high tumorgenicity (Yamashita et al., 2011). iPSCs are seeded on specific feeder layers, which aid the iPSC differentiation into e.g., DA neurons (Takahashi and Yamanaka, 2006;Stanslowsky et al., 2014) but direct neuronal conversion of fibroblasts into DA neurons in vitro and in vivo is possible with only three transcription factors: Asc11, Brn2 and Myt1l (Vierbuchen et al., 2010;Liu et al., 2012, Nakagawa et al., 2014. With the addition of two genes, Lmxa1 and FoxA2 (Pfisterer et al., 2011), the converted neurons can be further directed into a DA neuronal fate. This is confirmed by cell expression of the enzyme tyrosine hydroxylase (TH), which is involved in dopamine synthesis (Vermilyea et al., 2017). Fibroblasts can be substituted with peripheral blood cells (PBCs) and converted into DA neurons (Churko et al., 2013;Effenberg et al., 2015). This has its advantages, as it is easier, quicker, cheaper and less invasive to collect as opposed to a skin biopsy for the collection of fibroblasts (Zhou et al., 2015;Okumura et al., 2019). However, the blood cells are sensitive to temperature fluctuations (Baboo et al., 2019) and the patient might suffer from different clotting disorders (Kitchens et al., 2011), which can interfere with the quality of the PBCs. Furthermore, epigenetic markers might transfer from PBCs to the iPSCs and affect the pluripotency potential, that complicate the iPSCs differentiation (Polo et al., 2010). Epigenetic carryover is a general problem with iPSCs and gene reprogramming might be necessary prior to iPSCs application.

Gene reprogramming
Formerly, iPSC lines showed variability although they were derived from the same human donor due to the different genomic sites of integration of the retroviral vectors, which were used to express the reprogramming factors in clonal lines (Kilpinen et al., 2017). The first report of iPSCs revealed approximately 20 integrations per clonal line (Takahashi et al., 2007). However, replacements with non-integrating viral vectors such as adenovirus (Tashiro et al., 2010) or Sendai virus (Fusaki et al., 2009;Tarnawski et al., 2019) or other non-integrated methods e.g., recombinant proteins or Epstein-Barr nuclear antigen-1 (EBNA-1)-based episomal plasmids have been used successfully for reprogramming (Drozd et al., 2015). Additionally, pluripotent cells have been made transgene-free (Kaji et al., 2009). Genetic variations can occur during the reprogramming process itself (reprogramming-induced), in later stages in the prolonged culture (culture-induced) or they might be pre-existing in the parental somatic cells (Gore et al., 2011;Laurent et al., 2011). Additionally, the iPSCs increase growth rate due to adaption and signs of reduced apoptosis. The genetic variability of the clonal lines could influence the neuronal differentiation efficiency of iPSCs and result in inconsistency between disease lines and control lines (Carcamo-Orive et al., 2017;Kilpinen et al., 2017). The genetic variations produce phenotype variations, which are affected by e.g., point mutations, abnormal DNA-methylation including epigenetic memory and duplications with copy number variations (CNV) (DeBoever et al., 2017). The cell lines are karyotyped for chromosomal abnormalities and single nucleotide polymorphisms (SNPs). Small undetected CNVs in iPSC due to low resolution can be detected with comprehensive SNP arrays (Hussein et al., 2011). To address phenotypic variation issues, genome editing can e.g., introduce mutations in healthy iPSC lines or correct mutations in patient-derived iPSC lines (Soldner et al., 2011;Yusa et al., 2011) The application of iPSCs in PD Acta Neurobiol Exp 2020, 80 to create isogenic clonal lines, which only differ at the changed base(s). Novel, yet established mechanisms to induce these genetic modifications include zinc-finger nucleases, transcription-activator-like effector nucleases and clustered regularly interspaced short palindromic repeats-Cas9 (CRISPR-Cas9) . The CRISPR-Cas9 gene-editing mechanism (Jinek et al., 2012) can introduce specific disease-associated mutations, capacitate specific point homozygous and heterozygous mutations into iPSCs and edit a single gene copy as opposed to both copies (Paquet et al., 2016) to study precise combinations of disease-associated mutations. Furthermore, transgene-free murine leukemia virus-like particles loaded with Cas 9 and single-guide RNA ribonucleoproteins, called Nanoblades, can induce in vivo genome-editing. Additionally, Nanoblades have been complexed with donor DNA for a homology-directed repair or programmed with modified Cas9 to mediate transcriptional upregulation of specific genes (Mangeot et al., 2019). Following gene-editing, Cas9 resides in the cell and might cause unexpected off-target mutations in vivo (Schaefer et al., 2017). The gene-editing tool is inexpensive and easily accessible but scientific, medical and ethical considerations are necessary to ensure appropriate application (Baltimore et al., 2015).

Parkinson's disease
PD is a progressive neurodegenerative disorder that affects the motor and cognitive system. The aetiology of PD is unknown but the conceptualisation of PD continues to advance (Kalia and Lang, 2015). The disease is presumably caused by both environmental and genetic risk factors and 80% of patients are classified as having idiopathic PD (Simón-Sánchez et al., 2009;Biernacka et al., 2016). Neuropathologically, midbrain dopaminergic (mDA) neurons degenerate predominantly in substantia nigra pars compacta (SNc), which causes dopamine deficiency and leads to movement abnormalities including bradykinesia, resting tremor, rigidity and postural instability. As PD progresses, patients may develop cognitive impairment, depression and dementia (Hely et al., 2008;Pfeiffer et al., 2014). The mean age of PD diagnosis is 57 years and the disorder affects 1% of the population over 65 years of age (Lang and Lozano, 1998). The reason for the degeneration is unknown, but it presumably involves aggregation of the protein α-synuclein, that has undergone misfolding, and accumulates in neurons as Lewy Bodies (Spillantini et al., 1998). The accumulation of α-synuclein has also been detected in glia cells, which indicates that glia cells play a contributing non-cell autonomous role in PD neurodegeneration by inducing an inflamma-tory response (Wakabayashi et al., 2000). α-synuclein is encoded by the SNCA-gene (Devine et al., 2011). SN-CA-gene mutations as well as LRRK2-gene mutations are highly penetrant and can cause autosomal dominant PD (Kessler et al., 2018) and cause early onset PD (Nguyen et al., 2011). Other genes e.g., PARK2 and PINK1 presumably play a role in PD development (Kitada et al., 1998;Simón-Sánchez et al., 2009). The genes regulate key cellular processes and protein homeostasis including intracellular protein and membrane trafficking, neurite structure, ubiquitin-proteasome and lysosome-autophagy systems as well as mitochondrial function (Cuervo et al., 2004). Under physiological conditions, α-synuclein improves ATP synthase efficiency and membrane plasticity, however, when the protein aggregates into beta sheet-rich oligomers, a pathological gain-of-function is induced (Ludtmann et al., 2016). A study showed that iPSC-derived neurons with SNCA-triplication (SNCA-Tri) generated high levels of α-synuclein aggregates that interacted with the ATP-synthase and increased the probability of the permeability transition pore (PTP) opening leading to neuronal cell death (Ludtmann et al., 2018).
PD treatment is focused on normalizing the dopamine balance in the nigrostriatal system either by inhibiting the cholinergic fiber system, which dominates after the degeneration of the nigrostriatal projections, or, by improving and stimulating the activity of the remaining dopaminergic projections. The latter is currently the most effective and motor symptoms can be successfully reversed by re-establishing striatal dopaminergic neurotransmission with levodopa/carbidopa treatment (Lewitt, 2008). However, long-term levodopa use is associated with on-off fluctuations and hyperkinesia/dyskinesia (Borgkvist et al., 2018) and occur for 50% of the patients after years of levodopa use (Holloway, 2004). Neurosurgical interventions such as deep brain stimulation (DBS) can help maintain and extend the benefits of levodopa (Deuschl et al., 2006). DBS compensates for the electrical discharge caused by dopamine loss in the internal globus pallidus (GPi) or in the subthalamic nucleus (STN) (Saenger et al., 2017). Whilst current treatment might relieve symptoms, the effect is only temporary (Yoo et al., 2018).

The application of iPSCs
The application of iPSCs is valuable in many areas of disease research (Yusa et al., 2011;Kelaini et al., 2018;Zhao et al., 2019). Patient-derived iPSCs have enabled the study of patient-specific disease models (Trilck et al., 2016) which are useful for testing new therapeutic strategies and elucidating unknown pathology of e.g., neurodegenerative diseases. The disease models are particularly useful for the study of neurodegenerative diseases such as Huntington's Chorea (The Hd iPsc Consortium 2012), Alzheimer's disease (Shirotani et al., 2017) and PD (Kikuchi et al., 2017b), because the mechanisms behind the disorders are indistinct and only symptomatic treatment is available (Nakano and Tyler, 1971;Deuschl et al., 2006). Cell transplantation therapy is a possible regenerative treatment for neurodegenerative disorders and patient-specific iPSCs are promising for autologous cell transplantation therapy (Wang et al., 2015). Over the past decade, iPSC-technology has advanced with the development of novel techniques and reprogramming strategies for manipulating, modulating and differentiating iPSCs, which has given further insight into to the underlying pathology of neurodegenerative diseases such as PD Kim et al., 2019). The opportunities for iPSCs in PD are illustrated in Fig. 1 and addressed below.

Cellular disease modelling
The use of iPSCs as cellular disease models provide a unique possibility to investigate the underlying neuropathophysiology of PD . This is known as the 'disease in a dish' concept. Prior to cellular disease modelling, the identity and safety of the cell lines are verified by investigating their gene expressions. The European Bank for Induced Pluripotent Stem Cells has launched a catalogue consisting of standardized iPSCs for use in disease modelling in 2016 (De . Reprogrammed somatic cells derived from PD-patients can be differentiated into mDA neurons constituting formerly non-accessible in vitro models (Kaji et al., 2009). The reprogrammed human iPSCs (hiPSCs) are disease and patient-specific and possible gene mutations and/or chromosome abnormalities are retained during reprogramming, which is important for the study of genetic mutations in vitro (Reinhardt et al., 2013). Genetic mutations can interfere with key cellular processes and cause protein misfolding, lysosome dysfunction and oxidative stress (Flierl et al., 2014;Bogetofte et al., 2019). In a novel study (di Domenico et al., 2019), iPSC-derived astrocytes and DA neurons from PD patients with the LRRK2 G2019S mutation and healthy controls were generated in vitro. The control and patient-specific iPSC-derived DA neurons were co-cultured with either PD astrocytes or healthy astrocytes respectively. When the control DA neurons were co-cultured with the patient-specific iPSC-derived astrocytes, the axons and dendrites were shortened and disintegrated leading to neuronal death. Furthermore, astrocyte-derived α-synuclein accumulation was present. Conversely, healthy astrocytes co-cultured with DA neurons from PD patients, showed axonal and dendritic regeneration. Additionally, α-synuclein was prevented from accumulating, hereby restoring neuronal function. These results suggest the importance of astrocytes in PD pathogenesis and provide a fundament for exploring novel therapeutic strategies aiming to inhibiting the pathogenic crosstalk between DA neurons and astrocytes.
It has been hypothesized, that an imbalance in the ability of glia cells to detoxify increased production of reactive oxygen species (ROS) due to decreased glutathione levels and lack of neuroprotective growth factors such as glial-cell-derived neurotrophic factor (GDNF), may result in neuronal damage and eventually a progressive loss of DA neurons . Increased oxidative stress also exacerbates α-synuclein accumulation in vivo (Scudamore and Ciossek, 2018). Furthermore, age-related phenotypes in iPSCs can be induced in culture by exposure to mitochondrial toxins like rotenone or paraquat that cause uncontrolled oxidative stress (Nguyen et al., 2011;Ambasudhan et al., 2013). Likewise, the protein progerin induces early cellular senescence, which can facilitate ageing in vitro (Miller et al., 2013), which is key when modelling a late-onset age-related disease in culture with short preservation. In a recent study (Tagliafierro et al., 2019), hiPSC-derived DA neurons from a patient with SNCA-Tri and a healthy control were applied in a PD model in vitro using a novel method to intensify neuronal nuclear ageing by passing hiPSCs multiple times at the neural precursor cell stage, before the final differentiation. The patient-specific iPSC-derived DA neurons exhibited advanced nuclear neural ageing compared to the control DA neurons already at the Juvenile stage. The aged hiPSC-derived SNCA-Tri neurons contained more α-synuclein aggregates per cell compared to DA neurons at the Juvenile stage. These discoveries are relevant for future drug development and might also be relevant in the use of regenerative therapy to restore DA neurons.

Drug development
The 'disease in a dish' concept also provides in vitro models of PD that can be applied for novel drug screening. iPSC-technology provides an adequate number of cells necessary to complete the drug screening and it enables the investigation of several different cell types from the same PD patient (di Domenico et al., 2019). Hereby, iPSCs can be exploited to screen possible novel drugs and simultaneously screen for drug-induced toxicity in the cell cultures . The most promising drugs are selected for further in vivo studies to predict the feasibility of future therapeutics (Carrera et al., 2017;Romero et al., 2017). Therapies aiming at novel targets are needed to develop effective disease modifying treatments for PD (Hsieh et al., 2016). Much research has focused on enhancing the effects of levodopa (Johnston et al., 2018) but sporadic PD is a multifactorial disease and challenging to treat (Langston, 2006). Studies on genetic mutations associated with PD such as LRRK2 (Schapansky et al., 2018) and GAB1 (Sidransky and Lopez, 2012) have revealed interesting underlying mechanisms of PD pathogenesis. Furthermore, hiPSCs have been applied to screen and test experimental drugs Cao et al., 2016) in cooperation with pharmacological companies. Specifically, addition of the FDA-approved anti-IL-17 antibody to an iPSC-derived midbrain neuron culture from sporadic PD patients rescued DA neuronal death (Sommer et al., 2018). Ultimately, the validation and exploration of drug targets remain a bottle-neck in PD drug discovery. Relevant disease models, which recapitulate the complexity of PD to a greater extent, could further lead to improved drug discovery (Devine et al., 2011). Likewise, relevant animal disease models can be applied to test regenerative therapy such as cell therapy.

Regenerative therapy
The current effects of therapeutic strategies for PD have been incapable of stopping PD progression and restoring DA neural function (Nakano and Tyler, 1971) but cell therapy can potentially accomplish this (Wenning et al., 1997). PD is an interesting target for stem cell-based therapies due to the relatively focal degeneration of mDA neurons. For more than two decades, allografting of fetal ventral mesencephalic tissue has been conducted in clinical trials (Kordower et al., 1997;Li et al., 2016) with various outcomes, ranging from significant improvement to modest or no benefit at all (Lindvall et al., 1989(Lindvall et al., , 1990. This might be explained by lack of standardized fetal tissue, ethical issues associated with its use and post-operative incidences of dyskinesias in a number of patients (Freed et al., 2001;Olanow et al., 2003). iPSC-technology circumvents some of these issues but patient-specific iPSC production is expensive and additional safety issues regarding genetic and epigenetic variations as well as somatic cell use are associated with iPSCs (de Boni et al., 2018). With iPSC-technology, it is possible to use both human leukocyte antigen (Morizane et al., 2017) matched allogeneic (Shiba et al., 2016) and patient-specific autologous cell therapy (Mandai et al., 2017). Formerly, fetal grafts from tissue donors were not matched to the recipients, but the graft-derived dopaminergic innervation was maintained more than two decades after the transplantations (Li et al., 2016) and the indispensable immunosuppressive drugs were applied for less than two years or not at all (Freed et al., 2001;Olanow et al., 2003). Novel studies advocate the use of MHC-matched grafts because MHC-matching improves the engraftment of the iPSC-derived mDA neurons in animal models and reduces the immune response of microglia and lymphocytes. Although autologous transplantation is immunologically ideal, it is time consuming, expensive and not suited for standard therapy (Emborg et al., 2013;Hallett et al., 2015). Genetically, the autologous cells might also carry natural susceptibility to PD. Thus, allogeneic transplantation might prove to be superior to autologous transplantation (Morizane et al., 2017). Direct conversion of a somatic cell into an induced neuron (iN) circumvents safety concerns (Vierbuchen et al., 2010). The iNs are similar to iPSC-derived neurons and can be applied for HLA-matched and patient-specific treatments. Preclinical studies are still needed for the novel technique to elucidate the safety, stability and function of iNs prior to transplantation (Parmar, 2018). However, direct conversion of glia cells to iN derived from hiPSCs has increasingly gained attention. The method circumvents ethical concerns, risk of tumorigenesis and the need of transplantation to restore DA regeneration in PD (Matsuda et al., 2019).

Animal models
To further advance the preclinical studies, animal models that recapitulate PD better seems imperative as the predictive validity for clinical use and efficiency is based upon the resemblance between the animal models and PD patients (Kin et al., 2019). Ideally, the animal models should resemble human PD in terms of aetiology, pathology and behaviour and combine PD risk factors. Current models focus on aging (Miller et al., 2013), gene transfer (Reinhardt et al., 2013), the initiation of α-synuclein pathology by injection of Lewy Body extracts from PD patients (Recasens et al., 2014) and/or neurotoxins (Thomas et al., 2011) to mimic the disease as close to the human form as possible. Animal models intoxicated with the neurotoxins 1-methyl-4-phenyl-1, 2,3,6-tetrahydropyridine (MPTP) or 6-hydroxydopamine (6-OHDA) resemble a parkinsonian state with PD motor symptoms (Alvarez-Fischer et al., 2008;Huang et al., 2017). Reduction of the motor symptoms e.g., tremor and rigidity have been shown in animals after DA stem cell therapy (Kikuchi et al., 2011;Hallett et al., 2015). 6-OHDA and MPTP intoxicated animal models have been a staple in PD research for decades, but the models fail to recapitulate key features of PD such as the accumulation of misfolded α-synuclein. However, the neurotoxin β-sitosterol β-D-glucoside (BSSG) resemblances PD better with several important PD features such as Lewy body-like inclusions, late-stage cognitive effects and pre-motor changes e.g., olfaction in rat models (Van Kampen et al., 2015) and might be preferred when screening for neuroprotective approaches. The BSSG-model has been applied in several studies already (Van Kampen et al., 2014Wilson et al., 2002) and is undergoing an independent multi-site validation. Neurotoxins can be combined with genetic modification to mimic PD both genetically and biochemically (Thomas et al., 2011, Vila et al., 2001. Rodent models carrying genetic mutations e.g., LRRK2 (Nguyen et al., 2011) and PINK1 (Cooper et al., 2012) underwent patient-specific iPSC autologous cell transplantation but showed higher sensitivity to cellular stressors e.g., oxidative stress than fibroblasts with analogous mutations. However, idiopathic PD-specific iPSC (PDiPSC)-derived mDA neurons survived and functioned in 6-OHDA-lesioned rodent models, which suggests differences between the idiopathic and genetically modified animal models (Kikuchi et al., 2017b).
Cell cultures, rodent and large animal models contribute to deepen the comprehension of PD and the choice of model depends on the specific research question asked. Neurotoxin-lesioned non-human primates (NHPs) imitate humans genomically, neuroanatomically, physiologically and immunogenetically. They exhibit behavioural and symptomatic components of PD closer to the human condition as opposed to rodents, which have fast PD symptom development (Grow et al., 2016;Seo et al., 2019). However, the NHP models are expensive, raise an ethical concern and are therefore not universally allowed. Porcine models circumvent these issues Nielsen et al., 2016) and studies have shown promise providing Parkinsonian minipig models (Christensen et al., 2018;Lillethorup et al., 2018aLillethorup et al., , 2018b. Reduction in striatal DA neurotransmission and loss of TH-positive SN neurons as well as behavioural deficits and microglial activation have been shown. A possible use of the porcine model in regenerative therapy equivalent to the NHP model is presented in Fig. 2. However, porcine model studies lack longevity and further data on the PD pathology development is needed (Bjarkam et al., 2008).
Still, the NHP MPTP model often represents the last animal model of PD research used prior to human clinical trials. A preclinical study (Kikuchi et al., 2017a) using MPTP intoxicated NHPs (Macaca fascicularis) implanted with hiPSC-derived DA progenitor cells showed transplant survival after two years with no signs of implant oncogenesis. The progenitor cells functioned as mDA neurons and the mature mDA neurons projected fibers into the striatum of the recipient monkeys. 18F-DOPA PET scans showed increased uptake in the putamen following hiPSC transplantation compared to MPTP-treatment and the recipient monkeys restored spontaneous movements. However, a similar study using MHC-matched iPSC-derived neurons in a HD NHP model, failed to prevent graft rejection (Aron Badin et al., 2019). Furthermore, cell replacement therapy of DA neurons alone might be suboptimal as glia cells are important to PD pathogenesis and DA neuronal death as well (Gu et al., 2010;Zhang et al., 2011). Diseased astrocytes are unable to conduct neurotrophic andprotective functions, which leads to the death of their surrounding neurons. Transplantation of rodent astrocytes overexpressing e.g., GDNF via viral vectors in MPTP-intoxicated animal models have shown to induce behavioural recovery and neuroprotection (Drinkut et al., 2012). Likewise, a recent study showed that a viral delivery of GDNF improved graft outcomes of hiPSC-derived cell grafts in a rodent PD model (Gantner et al., 2020). Early exposure to GDNF endorsed survival and plasticity of non-DA neurons which improved the motor recovery in the PD model and delayed exposure to GDNF enhanced functional recovery via improved DA neuronal specification, plasticity and metabolism. Furthermore, it has been hypothesized, that oligodendrocytes and brain myelination play a role in PD pathogenesis (Dean et al., 2016). Studies of the shiverer mouse, which is a hypomyelinating mouse model, have showed myelin formation after transplantation with glial progenitor cells (GPCs) (Windrem et al., 2004(Windrem et al., , 2008. Furthermore, hiPSC-derived oligodendrocyte progenitor cells (OPCs) have been differentiated into both oligodendrocytes and astrocytes in vitro and in vivo. The hiP-SC-derived progenitor cells were engrafted into shiverer mice and myelinated the hypomyelinated mice successfully and increased their survival with no tumorigenesis nine months post-transplantation (Wang et al., 2013). Another study with the mouse model used both human and mouse glial restricted progenitors (GRPs), that were transplanted into the mice, however, only human GRPs migrated extensively and increased the life span of the dysmyelinated mice (Lyczek et al., 2017). This indicates that transplanted human GPCs might provide therapeutic benefits and that mature myelin might play an additional role in the CNS other than facilitating rapid neuronal conduction. Further studies of large animal models are needed to ensure the potential therapeutic effects in humans.

Clinical use
In 2014, research teams in Japan, USA and Europe created a global consortium, GForce-PD with the aim of cooperatively developing neuronal stem cell therapy for PD. Two clinical trials, CiRA trial and Summit for PD trial, were planned to apply iPSCs as initial cell source while two other human trials would apply hESCs (Fan et al., 2020). To our knowledge, one human clinical trial of allogeneic hiPSC-derived cells is currently ongoing. Based on the promising NHP results (Kikuchi et al., 2017a), the human clinical trial, CiRA, with iPSC-based therapy began in 2018 in Japan (Takahashi, 2019). Seven mid-stage PD patients with a L-DOPA response of > 30% Fig. 2. Parkinson's disease model. The hiPSCs are differentiated into progenitor DA neurons and can then be transplanted into the brain of any disease model subject e.g., the Göttingen minipig. Prior to transplantation, the Göttingen minipig has been intoxicated with MPTP or 6-OHDA to induce a Parkinson-like state. The functionality and safety of the transplantation can then be evaluated using behavioral tests, imaging techniques and immunohistochemical stains for e.g., TH-positive neuron reduction and microglial activation.
were selected for surgical transplantation of 4.8 million allogenic hiPSC-derived DA progenitor cells from an HLA-homozygous donor into the putamen bilaterally via stereotaxic brain surgery. The patients were treated with the immunosuppressant drug, Tacrolimus, to avoid rejection of the transplant. So far, no severe adverse events have been reported. The patients are to be followed for at least two years post-transplantation to comprehensively evaluate both qualitative and quantitative outcomes including motor, non-motor, cognitive, psychiatric and quality of life assessments. The graft function and survival, cell proliferation, microglia activation as well as a tumorigenesis are assed via MRI and/or [18F]DOPA-PET. The primary end points are tolerability, feasibility and safety with no adverse events. If the trials are successful, efficacy trials will be conducted. The secondary clinical end points are significant improvement of motor and non-motor assessments (Parmar et al., 2020). Additionally, California's Stem Cell Agency (CIRM) has received funding to conduct a clinical trial using autologous cell therapy for PD using patient-specific iPSC-derived DA neurons. The application of autologous stem cells circumvents the use of immunosuppressants (Loring, 2018). Furthermore, gene therapy will be applied to promote GD-NF-production to protect DA neurons. However, a novel study with intraputamenal GDNF to PD patients did not show significant clinical improvements against placebo in a randomized controlled trial after 40 weeks although it increased the [18F]DOPA uptake significantly in the putamen (Whone et al., 2019)

CONCLUSIONS
The invention of iPSC-technology has revolutionized the field of stem cell therapy in neurodegenerative disorders such as PD. The iPSCs have improved the possibilities to uncover the aetiology of PD and to identify novel targets for potential medical treatments (di Domenico et al., 2019;Laperle et al., 2020). Furthermore, the iPSCs are a promising source for cell transplantation as opposed to their embryonic counterparts because the iPSCs circumvent the ethical burden associated with the use of embryonic or fetal tissue (Freed et al., 2001;Takahashi and Yamanaka, 2006). The potential of iPSCs is immense, however, issues involved with the application of iPSCs still need to be addressed and resolved. The iPSC-technology should aim to become effective and safe in the reprogramming of the somatic cells in the future (Hu et al., 2010). Less comprehensive and standardized techniques and protocols for differentiation of iPSCs for production of isogenic cell lines are required (Nolbrant et al., 2017). Furthermore, the iPSCs have different origins, genetics, epigenetics and environmental factors, which have to be taken into consideration. Genomic malfunctions have been addressed but not yet perfected (DeBoever et al., 2017), just as full-genome sequencing and epigenomic analyses before application of iPSCs may be needed. As long as it is unknown whether the genetic changes have any functional significance, caution is of great importance particularly in the future clinical application. Although iPSC-derived therapies in PD have been promising in pre-clinical studies, the application of iPSC-technology is still in early stages in the clinical testing and uncertainties remains regarding the readiness for clinical testing and the following potential outcomes. The preclinical study on NHP implanted with hiPSC-derived DA progenitor cells was successful for the 2-year study period, however, other conducted studies with NHPs have not shown the same effect (Aron Badin et al., 2019). The long-term goal is to restore neuronal function in PD patients with hiPSCs. The first human clinical trials using hiPSCs have started in Japan (Takahashi, 2019) but the final results are yet to be published. PD is a complex disease with great individual variability in terms of pathology and progression. Perhaps, there is not one solution that fits all PD-patients and different approaches to treat PD including symptomatic treatment, neuroprotective regimens and cell therapy, may be needed. Nevertheless, it is plausible to think, that iPSC-based cell therapy will play a significant role in the future therapy for PD.