Use of human pluripotent stem cell-derived cells for neurodegenerative disease modeling and drug screening platform

Review For reprint orders, please contact: [email protected] Use of human pluripotent stem cell-derived cells for neurodegenerative disease modeling and drug screening platform Juan Antonio Garcia-Leon1,2, Javier Vitorica2,3,4 & Antonia Gutierrez*,1,2 1Departamento Biologia Celular, Genetica y Fisiologia, Facultad de Ciencias, Instituto de Investigacion Biomedica de Malaga-IBIMA, Universidad de Malaga, 29010 Malaga, Spain 2Centro de Investigacion Biomedica en Red Sobre Enfermedades Neurodegenerativas (CIBERNED), 28031 Madrid, Spain 3Departamento Bioquimica y Biologia Molecular, Facultad de Farmacia, Universidad de Sevilla, 41012 Sevilla, Spain 4Instituto de Biomedicina de Sevilla (IBiS)-Hospital Universitario Virgen del Rocio/CSIC/Universidad de Sevilla, 41013 Sevilla, Spain *Author for correspondence: +34 952133344; agutierr[email protected] Most neurodegenerative diseases are characterized by a complex and mostly still unresolved pathology. This fact, together with the lack of reliable disease models, has precluded the development of effective therapies counteracting the disease progression. The advent of human pluripotent stem cells has revolutionized the field allowing the generation of disease-relevant neural cell types that can be used for disease modeling, drug screening and, possibly, cell transplantation purposes. In this Review, we discuss the applications of human pluripotent stem cells, the development of efficient protocols for the derivation of the different neural cells and their applicability for robust in vitro disease modeling and drug screening platforms for most common neurodegenerative conditions. First draft submitted: 18 October 2018; Accepted for publication: 13 February 2019; Published online: 4 June 2019 Keywords: Alzheimer’s disease •amyotrophic lateral sclerosis •CRISPR-Cas9 •disease modeling •drug screening •hiPSCs •human-induced pluripotent stem cells •Huntington’s disease •multiple sclerosis •neurodegenerative disease •Parkinson’s disease Neurodegenerative diseases are characterized by the chronic and progressive deterioration of neuronal functioning, resulting in cognitive impairment, memory deficits, deficiency in motor function, loss of sensitivity, affectation of autonomous brain system, changes in perception and/or mood, etc., with combination of these features present in individual patients. Several neurodegenerative diseases share common pathological pathways including abnormal accumulation of toxic-aggregated proteins, mitochondrial dysfunction, axonal transport defects and chronic inflammation, which ultimately lead to neurodegeneration [1]. Neurodegenerative conditions represent an enormous impact on global health due to their high incidence, the severity of their symptoms and, in general, the lack of effective therapies able to mitigate those symptoms and/or counteract the course of the disease [2]. In addition, the incidence of neurodegenerative diseases in developed countries is rapidly increasing, as is the case for Alzheimer’s disease (AD) [3], amyotrophic lateral sclerosis (ALS) [4], multiple sclerosis (MS) [5], Parkinson’s disease (PD) [6] and other forms of dementia as frontotemporal dementia, progressive supranuclear palsy or corticobasal degeneration [7]. Although the reasons for this general increase in the number of people affected by neurodegenerative diseases are not properly established, the general aging of the western-countries population and a higher exposure to certain substances as chemicals and a higher contamination, are few of the possible causes for this increase [4]. Modeling of neurodegenerative diseases has always been a difficult task, due to the complexity of the conditions, poor access to human samples and little knowledge of disease pathology. Animal models have been the most widely used, due to the intrinsic features of in vivo experimentation which allow, in principle, more accurate modeling. Nonetheless, disease modeling with animals requires the generation of reliable disease models, the use of a large number of animals and limits the combination of experimental conditions tested, due also to the long experimentation times required [8]. Future Med. Chem. (2019) 11(11), 1305–1322 ISSN 1756-8919 130510.4155/fmc-2018-0520 C 2019 Antonia Gutierrez Review Garcia-Leon, Vitorica & Gutierrez The advent of human pluripotent stem cell (hPSC) discovery and development has represented a new scenario complementing the modeling of neurodegenerative diseases. Since their discovery, important efforts have been established in the derivation of hPSC-derived neural cells from patients and healthy donors, which allow disease modeling using human samples with inherent advantages over in vivo models. In this Review, we will discuss the current state of the art in neurodegenerative disease modeling and drug testing for the most prevalent neurodegenerative conditions, and provide data predicting the development of this fast-changing and promising field of research. Pluripotent stem cells Since the advent of human embryonic stem cells (hESCs) derivation in 1998 [9] and the discovery of murine and human-induced pluripotent stem cells (hiPSCs) in 2006–2007 [10,11], a new field of research was opened up. Although stem cells have represented a promising strategy for their potential therapeutic uses, which have already been investigated for more than two decades, successful research performed in the past few years have revealed a prominent utility of stem cells for disease modeling and drug screening purposes [12]. Human ESCs and iPSCs are pluripotent stem cells, with hESCs deriving from the inner mass of the blastocyst. On the other hand, hiPSCs are generated by the forced expression of reprogramming factors, with the transcription factors OCT3/4,SOX2,KLF4 and c-MYC constituting the initial reprogramming cocktail [11]. Both hESCs and hiPSCs have the capacity, under certain defined conditions, to differentiate toward any somatic cell type, as will be expanded later. The employment of hPSCs for disease modeling was initially delayed due to the relatively poor yield of initial methods for their derivation, complexity and costs. Nonetheless, several improvements in methodology for their generation have been developed. Among these improvements, combination of different reprogramming factors [13] and the combined use of different compounds [14] have contributed to increase the efficiency of reprogramming in a great extent. In addition, methods to introduce reprogramming factors relied in the use of integrating viral vectors as retrovirus and lentivirus due to their efficiency, although this technology precludes a potential further use of these cells and leads to random integration events hindering the standardization of this approach among labs. Recently, other methods for transgene delivery have been developed including the delivery of reprogramming factors by means of targeting the cells with episomal vectors, non-integrating Sendai virus and the use of mRNAs. These technologies lead to the generation of iPSCs with similar efficiencies that those derived using viral vectors and allow the generation of GMP-grade iPSCs compatible with their use in the clinical setting [15]. These advances have allowed the extension in the use of hPSCs among labs for disease modeling and drug screening purposes (Figure 1). This has represented a considerable advance in the field of neurodegenerative conditions that, until the implementation of hPSC technology, had been modeled exclusively and only partially by the use of animal models without full recreation of disease phenotype [16]. hPSC technology has transformed the modeling of neurodegenerative diseases, since it has allowed the study from familial monoallelic to complex multifactorial diseases in the context of patient-derived neural cells and tissue, systems that previously were not available for experimental investigation. The ability of hPSCs to differentiate toward any somatic cell type allows the generation of neural cells implicated in the pathogenesis of neurological diseases, where insertion or correction of disease-associated genetic mutations can be performed. Gene-editing in hPSCs An important and sometimes poorly considered challenge of hPSC technology is the variability between lines from the same individual or from the same group of individuals in terms of their capacity to differentiate toward different lineages. These differences are due to intrinsic genetic backgrounds, differences in reprogramming methods and/or epigenetic status of the somatic cell at the time of reprogramming [17]. Therefore, the generation of isogenic pair cell lines by correction or introduction of desired disease-associated genetic modification(s) is the best approach for assessing the biological effects exerted by one or several disease-associated mutations. Making such isogenic control lines requires the use of site-specific nucleases targeting the desired location of the genome. The use of site-specific nucleases to generate double-strand breaks in the genome allowed an efficient modification of the genetic sequence [18]. The complexity of the initial technologies and their scarce customization limited the wide implementation of this technology [19]. An important revolution in the field came by the employment of DNA-binding zinc finger nucleases for genomic editing, which notably increased the specificity and efficiency of the approaches, especially when cells were provided with a ‘donor plasmid’ (an exogenous DNA flanking the 1306 Future Med. Chem. (2019) 11(11) future science group Use of human pluripotent stem cell-derived cells for neurodegenerative disease modeling & drug screening platform Review Cell acquisition Patients iPSCs Disease-relevant neural cell type Somatic cells Reprogramming Pluripotent factors Differentiation to specifric neural cell type Cell transplantation Development of novel therapies Disease modeling and drug testing Figure 1. Schematic representation of the potential applications of induced pluripotent stem cells obtained from patients with neurodegenerative diseases. Human iPSCs can be differentiated toward disease-relevant neural cell types to be used potentially directly for cell transplantation or to be used as a platform for disease modeling and drug screening, which can lead to the development of novel and effective therapies. iPSC: Induced pluripotent stem cell. cutting site and with great homology with the target locus), which can be used by the cell as a template to repair the double-strand brake by homology recombination [20], a technology which can be applied to the genomic editing of hESC and hiPSCs as well [21]. In 2010, Christian et al. used a novel system by fusing transcription activator-like effectors (TALEs) to the catalytic domain of the Fok I endonuclease to generate TALE nucleases (TALENs) [22]. TALENs are easier to customize and less restrictive about the requirements the sequence must compile to be targeted. Moreover, this system showed a higher efficiency and accuracy over previous technologies [23]. However, the real revolution in the field came with the discovery and adaptation for genome editing of the CRISPR/CRISPR-associated (Cas) adaptive immune systems [24], with the variant CRISPR-Cas9 being the most widely used [25]. This complex is formed by the Cas9 nuclease associated with a base-paired structure formed between the activating tracrRNA and the targeting crRNA to cleave the target double-strand DNA. Site-specific cleavage occurs at places determined by both base-pairing complementarity between the CRISPR RNA (crRNA) and the target protospacer DNA and a short motif (the protospacer adjacent motif), next to the complementary region in the target DNA. As result, the Cas9 endonuclease can be programmed with single RNA molecules to future science group www.future-science.com 1307 Review Garcia-Leon, Vitorica & Gutierrez cleave specific DNA sites, leading to double-strand breaks in the genome that can be used for introducing desired modifications by co-providing a donor template DNA with desired modifications [24]. The advent of having a customizable RNA sequence (in contrast with protein sequence modifications needed for customization of zinc finger nucleases and TALENs) has resulted in a powerful gene-editing tool which, in addition, is cheaper and easier to design and implement than previous technologies. These features have made CRISPR-Cas9 system a widely used tool for gene-editing. Immediately after their discovery, they have been applied for performing gene-editing of hPSCs [26]. hPSCs are characterized by a poor efficiency of transfection, so specific gene-editing strategies should be followed to generate efficient modifications in the genome of these cells [27].In this sense, the incorporation of selectable cassettes in safe harbor loci in the genome [28] or a gene-editing strategy including the incorporation of selectable cassettes with a subsequent removal in a seamless manner [29] allow a more efficient targeting strategy. CRISPR-Cas9-mediated gene editing is an excellent tool for modeling some monogenic diseases or to study the contribution of single or several gene variants to a certain pathology. In this sense, hiPSCs derived from patients carrying a certain mutation associated with a neurodegenerative disorder can be used not only to model the disease but through the correction of this specific mutation and generation of isogenic cell lines, the contribution of a certain genetic modification to the disease phenotype can be addressed in a human model. In the case of familial ALS cases with mutations in the FUS gene, Guo et al. using iPSC-derived motor neurons (MNs) from these patients and correcting the mutations using CRISPR-Cas9 technology demonstrated that the presence of FUS mutations were responsible for shortcomings in axonal transport, a defect which could be counteracted by pharmacological inhibition of the HDAC6 gene as well, identifying a possible therapeutic target for the disease [30]. In the case of AD, studies employing the CRISPR-Cas9-mediated gene targeting in iPSCs are emerging, elucidating the pathogenic role of disease-associated mutations as those present in PSEN2 [31],PSEN1 [32],APP [33] or APOE [34].Moreover, CRISPR-Cas9-mediated gene editing has promised utility in the treatment of neurodegenerative diseases, as it has been shown for PD [35]. On the other hand, CRISPR-Cas9 can be used for the introduction of pathogenic variants for disease modeling. In this sense, we have recently generated a tauopathy model by introducing three MAPT mutations using CRISPR-Cas9 combined with the piggybac transposase in a seamless manner in iPSCs, leading to the manifestation of neurodegenerative phenotypes in iPSC-derived cortical neurons [36]. Therefore, gene-editing strategies have evolved in parallel with the use of hPSCs for disease modeling, a fact that has just started to prove fruitful for the combination of these two tools for the discovery and validation of novel therapeutic strategies [37]. Despite the potential of the combination of gene-editing in hPSCs for disease modeling and as a platform for drug testing, there are several drawbacks that limit this technology. One of these limitations deals with the fact that most neurodegenerative diseases are caused not by monogenic mendelian mutations but are consequence of environmental factors together with multiple low-effect-size-risk alleles identified in genome-wide association studies. This makes modeling with hPSCs a difficult task since disease-associated gene variants differ among patients and are not a causative agent per se to develop the disease or the pathogenic phenotype. To circumvent this issue, one possible strategy would be to generate several iPSC lines from patients with different genetic backgrounds in combination with multiplexed CRISPR-Cas9-mediate gene-editing to address the implication of those variants and/or discover novel disease-driving molecular pathways [38]. Another limitation of gene-editing strategies (with special focus on CRISPR-Cas9) deals with nuclease specificity and off-target events. Several studies have determined the rate of off-target events produced by site-specific nucleases, focusing on the CRISPR-Cas9 system. Initial studies determined that the possibility of having off-target effects in a targeted cell or organism was ≥50% [39], limiting the use of this technology. Nonetheless, more accurate methods for measuring those off-target events [40] and the last optimizations of the CRISPR-Cas9 technology have reduced practically to undetectable the frequency of off-target mutations [41]. However, this phenomenon should always be taken into account for the experimentation performed. Methods for the derivation of neural cells from hPSCs One of the intrinsic features of pluripotent stem cells is that, when exposed to defined culture conditions, are able to differentiate toward (almost) any somatic cell type [42]. Based on this, soon after the derivation of hESCs, and especially after the discovery of hiPSCs, researchers have made important efforts to elucidate molecular pathways governing cell differentiation toward the different lineages, with a special focus on the derivation of neural cells, since access to primary brain tissue is scarce and, when available, expansion of the different cell types ex vivo is 1308 Future Med. Chem. (2019) 11(11) future science group Use of human pluripotent stem cell-derived cells for neurodegenerative disease modeling & drug screening platform Review Diseases ADCortical neurons PSCs Neural induction Neuronal maturation N2 B27 DA neuron maturation GABAergic maturation Pur, N2B27 Astrocyte induction Heregulin 1b, IGF-1, FBS MN induction -SMAD, -TGFb, RA, SAG Shh, FGF8, BDNF, AA, GDNF, cAMP -SMAD/-TGFb NPCs MNP MN maturation DAPT, CNTF OL maturation OPC induction Microglia induction MCSF, IL-34, TGFb1, insulin -SMAD, -TGFb, RA, SAG Microglia maturation CD200, CX3CL1 T3, AA, PDGF, HGF, IGF-1 NT3, biotin, cAMP OPCs MP DA neurons GABAergic neurons Astrocytes Motor neurons Oligodendrocytes Microglia Brain organoid Organoid induction PD HD ALS MS, ALS AD, MS, ALS AD, MS, ALS AD, MS, ALS Figure 2. Layout of culture conditions for the derivation of human pluripotent stem cells into the different neural cell types and relevance of each cell type to neurodegenerative diseases. AD: Alzheimer’s disease; ALS: Amyotrophic lateral sclerosis; HD: Huntington’s disease; MS: Multiple sclerosis; PD: Parkinson’s disease. very limited. Next, we are going to discuss the progress made for the optimization of the differentiation methods developed to generate the different neural cell types from hPSCs (see Figure 2). Induction of a neural fate & neuronal differentiation Soon after hESCs derivation, different attempts were made to fate pluripotent stem cells toward a neural phenotype. It was observed that when hPSCs are cultured in standard conditions without pluripotency-sustaining factors, neural fate is spontaneous induced, with the TGFβ-related signaling negatively inducing this differentiation [43]. Later on, using specific TGFβantagonists to abolish the SMAD signaling, Chambers et al. improved and defined future science group www.future-science.com 1309 Review Garcia-Leon, Vitorica & Gutierrez the differentiation method obtaining a robust and efficient neural conversion from both hESCs and hiPSCs [44]. Nonetheless, these derivation methods did not recreate human neurogenesis in a temporal manner, with important limitations for disease modeling. In 2012, Shi et al. circumvented these issues since they developed an improved method for neuronal specification based on known rodent cortical development that included, after neural progenitor specification based on TGFβ-signaling inhibitors, an extended period of cortical neurogenesis and terminal neuronal maturation, which recreate human corticogenesis with the generation of deep and upper layer cortical neurons in a specified temporal manner [45]. This methodology enables studies to unravel the development of human cerebral cortex and the generation of individual-specific neurons that could be used for disease modeling and therapeutic approaches. Further improvements in the methodology to obtain electrophysiological active neurons have been described [46], also using defined and standard culture conditions without the need of co-culturing the generated neurons with exogenous cells [47]. In addition, protocols for generating specific neuronal subtypes have been established. In this sense, specific protocols have been developed to generate dopaminergic (DA) neurons from hPSCs. Nguyen et al. achieved efficient generation of DA neurons by a protocol performing neural induction by dual-SMAD inhibition, followed by culturing the cells in specific DA neurons cues including Sonic hedgehog and FGF8b signaling, supporting the cells with additional growth factors as BDNF, ascorbic acid, GDNF and cAMP; obtaining DA neurons able to reproduce key PD features in vitro [48]. Further optimizations of the protocol have been performed, leading to a more efficient generation of DA neurons able to engraft murine and non-human primates [49–51]. MNs have been another neuronal population widely studied due to their implication in several neurodegenerative conditions as ALS, where selective MN death occurs [52]. Since the discovery of hiPSCs, several protocols have been developed for the specific differentiation of these cells into functional MNs [53–56].Mauryet al. tested several small molecules targeting developmental cues for the derivation of different neuronal subtypes [57]. They obtained an improved method for MN generation consisting of neural specification induced by SMAD inhibition and Wnt pathway activation, followed by caudalization and ventralization induced by retinoic acid and smoothened agonist, respectively; culturing the cells in the presence of the gamma secretase inhibitor N-[N-(3,5-Difluorophenacetyl)- L-alanyl]-S-phenylglycine t-butyl ester (DAPT) and ciliary neurotrophic factor [30]. Although cell-replacement therapies are far from being achieved due to the complex and still unsolved pathology of ALS, hPSC-derived MNs can be used to resolve the complex pathology and serve as a platform to test altered pathways [37]. Apart from DA or MNs, other neuronal subtypes have been generated by finely exposing pluripotent cells to the signals governing telencephalic development, in a time and concentration-dependent manner [58]. Generation of astrocytes from hPSCs Astrocytes are the most common glial cells present in the CNS, with essential functions in neuronal metabolic support, synapse formation and regulation of blood flow, among others [59]. Astrocytes are a heterogeneous cell type emerging from neural precursors in common with neurons that results in mixed generation of neurons and astrocytes in several protocols of hPSC differentiation [47]. Nonetheless, specific methods have been developed for efficient generation of functional astrocytes from hPSCs in defined culture conditions that resemble primary astrocytes and maintain their functionality after in vivo transplantation [60]. Further optimization in the protocols has been performed, obtaining more efficient methods to generate astrocytes able to unleash immune responses and able to support phagocytosis mediated by microglial cells [61]. hPSC-derived astrocytes can be generated from individuals suffering from different neurodegenerative diseases and can reproduce several phenotypes encountered in vivo,a feature that can be employed for effective disease modeling and the discovery of new therapeutic targets [62–64]. Generation of oligodendrocytes from hPSCs Since the successful isolation and culture of hESCs, different groups developed specific protocols for the generation of oligodendrocyte precursor cells (OPCs) and oligodendrocytes (OLs), which are the cells responsible for myelin synthesis in the CNS. OLs and OPCs are a heterogeneous population of cells existing at different maturation states that, apart from being the cells responsible for myelination, have important trophic and metabolic functions [65]. Nistor et al. in 2005 was the first to describe the generation of OLs from hPSCs through neurosphere formation after a 10–15 week protocol, using only one hESC line [66]. Further improvements in the methodology were made, but all these protocols remained inefficient (very long differentiation times and low yield) and their robustness was in debate due to reduce number of cell lines tested [67,68]. All these protocols attempted to recapitulate the different processes and signals that occur during development. However, the different protocols differed in the combinations 1310 Future Med. Chem. (2019) 11(11) future science group Use of human pluripotent stem cell-derived cells for neurodegenerative disease modeling & drug screening platform Review of growth factors, cytokines and other molecules needed, which highlight that OL specification in vivo is achieved in different developmental regions that respond to specific local concentration of different factors [69]. Since the discovery of iPSCs, several groups developed specific protocols to obtain OPC/OLs from hiPSCs. Initial studies adapted previous protocols developed for hESCs to hiPSCs [70,71]. In 2013, Wang et al. developed an optimized protocol for OPC/OL generation from hiPSCs, accounting for a total differentiation time of 150 days, time after which they obtained 70–80% of OPCs, with a small fraction of those differentiating toward mature MBP+OLs able to engraft and myelinate the myelin-deficient shiverer mice [72]. Douvaras et al. in 2014 published an improved version of the protocol, consisting in a first stage generating OLIG2+OPCs in suspension, which were dissociated and plated to mature till day 75–90, when MBP+mature OLs started to appear [73]. Generated OPCs/OLs using this protocol were able to myelinate neurons in vivo and could be generated from MS patient lines. In line with this, Piao et al. showed the clinical applicability of hPSC derived-OPCs that were able to remyelinate and rescue an irradiated rat model of brain tumor radiation [74]. Although the last protocols reduced the required time needed for OPC/OL generation and increased the efficiency, the generation of OLs from hPSCs remained inefficient and required long differentiation times in culture, limiting the use of these cells. To circumvent these issues, we and others have developed specific protocols that notably increase the efficiency of OL generation and reduced the differentiation time [75,76]. We demonstrated that forced expression of SOX10 alone leads to efficient OL generation within 20 days from hPSC stage, with generated OLs able to myelinate neurons both in in vivo and in vitro contexts. Moreover, hPSC-derived neuron-OL cocultures can be used for addressing the effect on myelin production of myelinating drugs in a high-throughput manner, serving this system as a platform where to test and/or validate the effects of candidate drugs [76]. Generation of microglia from hPSCs The developmental origin of microglia has recently been elucidated, with few studies demonstrating that microglia arise from early yolk sac-derived hematopoietic progenitors, which migrate to the brain early in the development [77]. This fact has delayed the development of specific protocols to generate microglial cells from hPSCs. Several protocols were developed generating CD14+monocytes/macrophages from hPSCs [78,79]. These cells express phagocytic markers, are morphologically similar to blood-derived monocytes and macrophages, produce cytokines in a similar fashion as macrophages depending on the activation status of the cell and display high phagocytic capacity [80]. Other protocols tried to generate specifically microglial cells, but these protocols failed to resemble bona fide microglia since they did not recapitulate precisely microglial ontogeny [81,82]. In the past 2 years, a number of different protocols for generating microglial-like cells from hPSCs have been developed [83–86]. Although the methodology differed among the different approaches developed, these protocols have in common a first phase where early mesodermal progenitors are induced from hPSCs, followed by a maturation state where microglial-specific signaling pathways as IL34 and TGFb are induced [83–86]. This results in the generation of microglial-like cells resembling primary microglia at the transcriptome level, able to response to immune inducers and able to phagocytose amyloid-β, serving these cells as a platform for disease modeling and possibly drug testing [86]. Brain organoids Much of our current understanding of human brain development and functioning relied on experimental studies performed mostly on mouse models, which do not fully mimic the features present in the human brain. Threedimentional CNS organoids possess the unique advantage of recapitulating multiregional and region-specific cytoarchitecture seen in the early human fetal brain development, becoming a strong complement for studying brain development and pathology, to develop new therapies for treating neurodevelopmental diseases [87].Inthe past few years, different groups have reported the generation of human brain organoids from PSCs. Lancaster et al. were the first to demonstrate that brain organoids from hPSCs can be generated that recapitulate the human brain development generating different discrete but connected brain areas defined by the expression of specific cell markers and transcription factors, which can be used as well for modeling microcephaly [88]. After this discovery, other groups have used this approach to generate brain organoids that have resulted useful for modeling different developmental diseases as autism spectrum disorders [89], schizophrenia [90], microcephaly induced by Zika virus infection [91] and others [87]. In addition, 3D culture systems have been proved useful for modeling neurodegenerative diseases as AD [92,93] and dys-/demyelinating diseases [94]. future science group www.future-science.com 1311 Review Garcia-Leon, Vitorica & Gutierrez Use of hPSC-derived cells for drug screening & validation platform For decades, drug discovery and validation relied on experimental testing of a certain number of compounds in animals to observe the effects produced on phenotypes, processes that were costly and time-consuming, allowing the test of a limited number of compounds. Advances made in chemical synthesis and automation platforms altered the paradigm of drug discovery, allowing the development of high-throughput screenings (HTS). HTS allow testing of hundreds or even thousands of compounds in a short period of time and with relative low costs. Classical HTS have been based on simple platforms where the effect of candidate compounds were tested on purified proteins, RNA, DNA or a simple mixture of a few molecules [95]. More complex platforms relied on the use of simple organisms (as yeasts) or cell lines that allow to study the effect of candidate drugs on alive organisms [96]. Although efficient, these assays usually present a major problem that is the poor validation rates of initially identified hits in in vivo models reproducing the disease, due mostly to the scarce biological relevance of the screening platforms related to the disease to treat and the use of non-human derivatives, which do not hold specific human biology. This problem may be circumvented by the use of more reliable models as screening platforms. In this sense, hPSCs may represent an excellent system where to perform drug validation and even being used for drug discovery. Although inefficient protocols for the derivation of specific cell types have precluded the use of this technology for drug testing [65], we believe recent literature indicates that it is the right moment to perform this technology translation, and several small biotech companies as well as big pharma industries are developing stem cell-derived platforms for drug testing. Certainly, the use of hPSCs for drug testing has been earlier implemented in the case of hPSC-derived hepatocytes [97] and cardiomyocytes [98], possibly because the phenotypes to be observed after drug treatment are robust and easy to evaluate. For neurodegenerative diseases, this implementation has been slower due mostly to the inherent complex disease pathology, the inefficient protocols for the generation of neural cells and the lack of clear phenotypes to be evaluated. Next, we are going to discuss recent advances in disease modeling and drug screening for the main neurodegenerative diseases focusing in the employment of hiPSCs. Alzheimer’s disease AD is the most prevalent neurodegenerative disease and form of dementia, characterized by the accumulation of amyloid-beta (Ab) plaques and TAU-laden neurofibrillary tangles, with huge impact on society [99]. The pathology of AD is very complex and still not fully understood, with a recently described key role for microglial dysfunction in the disease [100–102]. Due to this fact and because it is a noncell autonomous disease, the disease modeling with hPSC-derived cells in AD has been hindered. Most of the studies performed have focus on reproducing the Ab pathology present in AD (aberrant extracellular accumulation of the Ab42 peptide). Several groups have demonstrated that neurons generated from iPSCs of AD patients show an increased secretion of Ab42 [103] and this platform can be exploited for testing of drugs diminishing this phenomenon. Yahata et al. in 2011 generated hiPSC-derived cortical neurons, able to produce Ab40 and 42 peptides. They inhibited the function of beta and gamma secretases (enzymes implicated in the synthesis of Ab) by employing beta- and gamma-secretase drug inhibitors as well as a nonsteroidal anti-inflammatory drug, finding that those drugs reduced Ab40–42 content [104]. Using a similar platform, Xu et al. tested over 100 drugs, finding that a Cdk2 inhibitor was able to reduce Ab toxicity [105]. Nonetheless, drugs targeting Ab production pathways have resulted inefficient in clinical trials [106], evidencing a more complex disease pathology. In this sense, Mertens et al. found that iPSC-derived neurons do not respond to clinical doses of nonsteroidal anti-inflammatory drug-based gamma-secretase modulation [107].Kondoet al. established an iPSC-derived neuronal drug screening platform where they tested over 1000 compounds for their ability of reducing Ab load within the culture. They obtained 27 hits that resulted in six leading compounds and, to maximize the anti-Ab effect, a combination of three compounds (bromocriptine, cromolyn and topiramate) was selected as anti-Ab cocktail, which showed efficacy in reducing Ab levels in iPSC-derived neurons, suggesting that this approach could be translated to clinical trials [108]. Israel et al. demonstrated that neurons derived from iPSCs of both sporadic and familial forms of AD behave differently, with not only anomalous Ab expression but also aberrant expression of the aGSK-3b and hyperphosphorylated TAU forms, linking both Ab and TAU pathological hallmarks in iPSC models. In addition, treatment with b-secretase inhibitors led to a reduction in phosphorylated TAU and aGSK-3b levels [109]. TAU pathology is present in AD and other dementias, and iPSC-derived neurons with mutations in the MAPT gene showed altered TAU isoform expression, hyperphosphorylated TAU aggregates as well as different phenotypes associated with neurodegeneration [110]. Recently, we generated an iPSC-derived mutant TAU model that shows several phenotypes 1312 Future Med. Chem. (2019) 11(11) future science group Use of human pluripotent stem cell-derived cells for neurodegenerative disease modeling & drug screening platform Review associated with neurodegeneration, among them the accumulation of pathogenic TAU. The characteristics of this iPSC-derived model allow the use of this platform for drug screening purposes [36]. As mentioned before, AD is a noncell autonomous disease, with involvement of several neural cell types. In fact, iPSC-derived astrocytes show pathological phenotypes in those derived from AD patients in comparison with those derived from healthy subjects [62,63], although still this platform has not been translated into the drug discovery field. As mentioned earlier, only recently specific protocols for the generation of iPSC-derived microglial cells have been developed [83–86], which have impeded modeling the role that microglia has in AD pathology. iPSC-derived microglia express main immunological risk variants associated to sporadic AD and are activated after Ab treatment [86]. Moreover, iPSC-derived microglia are starting to be used as a model where to test the effect of genetic variants in the microglial functionality [111,112], which can lead to the discovery of novel disease-mediated mechanisms. Finally, due to the involvement of several neural cell types in AD pathology, few studies have reported the generation of 3D organoid cultures that do not only reproduce AD pathology, but are also sensitive to drugs reverting disease phenotype [92,113]. Nonetheless, these approaches are far for being translated to HTS for drug discovery. Parkinson’s disease PD symptoms are caused by a specific loss of DA neurons in the ventral midbrain [114]. Due to this localized neuron loss, important efforts have been made to generate and evaluate the correct neural population to be transplanted into the putamen of assessed models. The use of neural cells from human origin for therapeutic purposes in terms of promoting recovery after injury or for cell replacement purposes has been long tested in PD, due to the specific neuronal population and region to be targeted. Already in the early 1980s, researchers demonstrated that transplantation of fetal DA neurons successfully engrafted and restored functionality in a PD rat model [115].This and other studies paved the way for the translation of this approach to patients 10-years later, when fetal DA neurons were successfully transplanted in a severe PD case, which engrafted and improved motor functions [116]. Since then, numerous studies have been performed employing neural cells from different origins [117]. Among them, the transplantation of DA neurons derived from iPSCs is taking shape to become a therapeutic possibility for PD patients. In this sense, recent protocols for obtaining DA neurons from hPSCs have notably improved the efficiency of obtaining these cells [118]. Preclinical studies have demonstrated that generated DA neurons, after purification, are successfully engrafted in a PD rat model [49]. The clinical suitability of hPSC-derived DA neurons has been further demonstrated, since these cells were transplanted in nonhuman primates treated with the neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), engrafting and improving motor function, surviving for at least 2 years without evidences of tumor formation [51]. This therapeutic approach is currently being translated to PD patients in a Phase I clinical trial. Several groups have generated DA from iPSCs of PD patients and healthy controls and compared altered pathways between them with the intention of use these cells for disease modeling (reviewed in [118]). When compared with healthy controls or PD patients from different genetic backgrounds, different altered pathways were identified, with some of them susceptible of therapeutic modulation. Hartfield et al. described the generation of mature and physiologically active DA neurons, which not only formed functional synapses but were susceptible to the DA-specific toxin 1-methyl-4-phenylpyridinium (MPP+) that reduced mitochondrial membrane potential and altered mitochondrial morphology, suggesting the usefulness of this system for the testing of drugs [119]. Moreover, iPSC-derived DA neurons from PD patients with familial forms of the disease are characterized for presenting several pathogenic phenotypes, suggesting that a combination therapy may be necessary for PD, employing patient-derived cells for its elucidation [120]. Therefore, for the development of effective therapies for PD, iPSC-derived platforms seem suitable to test candidate drugs, which will most likely result in a combination of cell and drug therapy for the treatment of PD symptoms and disease course. Huntington’s disease Huntington’s disease (HD) is an inherited neurodegenerative disorder caused by an expanded stretch of CAG trinucleotide repeats in the huntingtin gene that results in the loss of GABAergic neurons in the striatum [121]. iPSC lines from HD patients have been generated that after differentiation toward NSCs or neurons exhibited decreased cell adhesion and ATP production, increased apoptosis and cell death after prolonged culture or BDNF withdrawal, and increased vulnerability to stressors. Moreover, the severity of these disease-associated phenotypes are directly influenced by the extension of CAG repeats [122]. These phenotypes can be reversed after specific future science group www.future-science.com 1313 Review Garcia-Leon, Vitorica & Gutierrez 86. Abud EM, Ramirez RN, Martinez ES et al. iPSC-derived human microglia-like cells to study neurological diseases. Neuron 94(2), 278–293.e9 (2017). 87. Wang H. Modeling neurological diseases with human brain organoids. Front. Synaptic Neurosci. 10, 15 (2018). 88. Lancaster MA, Renner M, Martin C-A et al. Cerebral organoids model human brain development and microcephaly. Nature 501(7467), 373–379 (2013). •First demonstration of the generation of hiPSC-derived brain organoids. 89. Mariani J, Coppola G, Zhang P et al. FOXG1-dependent dysregulation of GABA/glutamate neuron differentiation in autism spectrum disorders. Cell 162(2), 375–390 (2015). 90. Stachowiak EK, Benson CA, Narla ST et al. Cerebral organoids reveal early cortical maldevelopment in schizophrenia – computational anatomy and genomics, role of FGFR1. Transl. Psychiatry 7(11), 6 (2017). 91. Dang J, Tiwari SK, Lichinchi G et al. Zika virus depletes neural progenitors in human cerebral organoids through activation of the innate immune receptor TLR3. Cell Stem Cell. 19(2), 258–265 (2016). 92. Choi SH, Kim YH, Hebisch M et al. A three-dimensional human neural cell culture model of Alzheimer’s disease. Nature 515(7526), 274–278 (2014). •Modeling Alzheimer’s disease in a 3D iPSC-derived model. 93. Raja WK, Mungenast AE, Lin Y-T et al. Self-organizing 3D human neural tissue derived from induced pluripotent stem cells recapitulate Alzheimer’s disease phenotypes. PLoS ONE 11(9), e0161969 (2016). 94. Madhavan M, Nevin ZS, Shick HE et al. Induction of myelinating oligodendrocytes in human cortical spheroids. Nat. Methods 15(9), 700–706 (2018). 95. Varma H, Lo D, Stockwell B. High throughput screening for neurodegeneration and complex disease phenotypes. Comb. Chem. High Throughput Screen. 11(3), 238–248 (2008). 96. Karapetyan YE, Sferrazza GF, Zhou M et al. Unique drug screening approach for prion diseases identifies tacrolimus and astemizole as antiprion agents. Proc. Natl Acad. Sci. 110(17), 7044–7049 (2013). 97. Cayo MA, Mallanna SK, Di Furio F et al. A drug screen using human iPSC-derived hepatocyte-like cells reveals cardiac glycosides as a potential treatment for hypercholesterolemia. Stem Cell 20, 478–489.e5 (2017). 98. del ´ Alamo JC, Lemons D, Serrano R et al. High throughput physiological screening of iPSC-derived cardiomyocytes for drug development. Biochim. Biophys. Acta 1863(7), 1717–1727 (2016). 99. Polanco JC, Li C, Bodea L-G, Martinez-Marmol R, Meunier FA, G¨ otz J. Amyloid-βand tau complexity – towards improved biomarkers and targeted therapies. Nat. Rev. Neurol. 14(1), 22–39 (2017). 100. Gutierrez A, Vitorica J. Toward a new concept of Alzheimer’s disease models: a perspective from neuroinflammation. J. Alzheimer’s Dis. 64(Suppl. 1), S329–S338 (2018). 101. Navarro V, Sanchez-Mejias E, Jimenez S et al. Microglia in Alzheimer’s disease: activated, dysfunctional or degenerative. Front. Aging Neurosci. 10, 140 (2018). 102. Sanchez-Mejias E, Navarro V, Jimenez S et al. Soluble phospho-tau from Alzheimer’s disease hippocampus drives microglial degeneration. Acta Neuropathol. 132(6), 897–916 (2016). 103. Yagi T, Ito D, Okada Y et al. Modeling familial Alzheimer’s disease with induced pluripotent stem cells. Hum. Mol. Genet. 20(23), 4530–4539 (2011). 104. Yahata N, Asai M, Kitaoka S et al. Anti-Aβdrug screening platform using human iPS cell-derived neurons for the treatment of Alzheimer’s disease. PLoS ONE 6(9), e25788 (2011). 105. Xu X, Lei Y, Luo J et al. Prevention of β-amyloid induced toxicity in human iPS cell-derived neurons by inhibition of cyclin-dependent kinases and associated cell cycle events. Stem Cell Res. 10(2), 213–227 (2013). 106. Winblad B, Amouyel P, Andrieu S et al. Defeating Alzheimer’s disease and other dementias: a priority for European science and society. Lancet Neurol. 15(5), 455–532 (2016). 107. Mertens J, St¨ uber K, Wunderlich P et al. APP processing in human pluripotent stem cell-derived neurons is resistant to NSAID-based γ-secretase modulation. Stem Cell Rep. 1(6), 491–498 (2013). 108. Kondo T, Imamura K, Funayama M et al. iPSC-based compound screening and in vitro trials identify a synergistic anti-amyloid β combination for Alzheimer’s disease. Cell Rep. 21(8), 2304–2312 (2017). 109. Israel MA, Yuan SH, Bardy C et al. Probing sporadic and familial Alzheimer’s disease using induced pluripotent stem cells. Nature 482(7384), 216–220 (2012). 110. Iovino M, Agathou S, Gonz´ alez-Rueda A et al. Early maturation and distinct tau pathology in induced pluripotent stem cell-derived neurons from patients with MAPT mutations. Brain 138(11), 3345–3359 (2015). 111. Garcia-Reitboeck P, Phillips A, Piers TM et al. Human induced pluripotent stem cell-derived microglia-like cells harboring TREM2 missense mutations show specific deficits in phagocytosis. Cell Rep. 24(9), 2300–2311 (2018). 1320 Future Med. Chem. (2019) 11(11) future science group Use of human pluripotent stem cell-derived cells for neurodegenerative disease modeling & drug screening platform Review 112. Brownjohn PW, Smith J, Solanki R et al. Functional studies of missense TREM2 mutations in human stem cell-derived microglia. Stem Cell Rep. 10, 1–14 (2018). 113. J Siney E, Kurbatskaya K, Chatterjee S, Prasannan P, Mudher A, Willaime-Morawek S. Modelling neurodegenerative diseases in vitro: recent advances in 3D iPSC technologies. AIMS Cell Tissue Eng. 2(1), 1–23 (2018). 114. Poewe W, Seppi K, Tanner CM et al. Parkinson disease. Nat. Rev. Dis. Prim. 3, 17013 (2017). 115. Perlow MJ, Freed WJ, Hoffer BJ, Seiger A, Olson L, Wyatt RJ. Brain grafts reduce motor abnormalities produced by destruction of nigrostriatal dopamine system. Science 204(4393), 643–647 (1979). 116. Lindvall O, Brundin P, Widner H et al. Grafts of fetal dopamine neurons survive and improve motor function in Parkinson’s disease. Science 247(4942), 574–577 (1990). 117. Yasuhara T, Kameda M, Sasaki T, Tajiri N, Date I. Cell therapy for Parkinson’s disease. Cell Transplant. 26(9), 1551–1559 (2017). 118. Li H, Jiang H, Zhang B, Feng J. Modeling Parkinson’s disease using patient-specific induced pluripotent stem cells. J. Parkinsons Dis. 8(4), 479–493 (2018). 119. Hartfield EM, Yamasaki-Mann M, Ribeiro Fernandes HJ et al. Physiological characterisation of human iPS-derived dopaminergic neurons. PLoS ONE 9(2), e87388 (2014). 120. Woodard CM, Campos BA, Kuo S-H et al. iPSC-derived dopamine neurons reveal differences between monozygotic twins discordant for Parkinson’s disease. Cell Rep. 9(4), 1173–1182 (2014). 121. Bates GP, Dorsey R, Gusella JF et al. Huntington disease. Nat. Rev. Dis. Prim. 1, 15005 (2015). 122. HD iPSC Consortium. Induced pluripotent stem cells from patients with Huntington’s disease show CAG-repeat-expansion-associated phenotypes. Cell Stem Cell. 11(2), 264–278 (2012). 123. Xu X, Tay Y, Sim B et al. Reversal of phenotypic abnormalities by CRISPR/Cas9-mediated gene correction in huntington disease patient-derived induced pluripotent stem cells. Stem Cell Rep. 8(3), 619–633 (2017). 124. Zhang N, Bailus BJ, Ring KL, Ellerby LM. IPSC-based drug screening for Huntingtons disease. Brain Res. 1638, 42–56 (2016). 125. Zuccato C, Valenza M, Cattaneo E. Molecular mechanisms and potential therapeutical targets in Huntington’s disease. Physiol. Rev. 90(3), 905–981 (2010). 126. Sarantos MR, Papanikolaou T, Ellerby LM, Hughes RE. Pizotifen activates ERK and provides neuroprotection in vitro and in vivo in models of Huntington’s disease. J. Huntingtons. Dis. 1(2), 195–210 (2012). 127. Nekrasov ED, Vigont VA, Klyushnikov SA et al. Manifestation of Huntington’s disease pathology in human induced pluripotent stem cell-derived neurons. Mol. Neurodegener. 11(1), 27 (2016). 128. Taylor JP, Brown RH, Cleveland DW. Decoding ALS: from genes to mechanism. Nature 539(7628), 197–206 (2016). 129. Hu B-Y, Zhang S-C. Differentiation of spinal motor neurons from pluripotent human stem cells. Nat. Protoc. 4(9), 1295–1304 (2009). 130. Sances S, Bruijn LI, Chandran S et al. Modeling ALS with motor neurons derived from human induced pluripotent stem cells. Nat. Neurosci. 19(4), 542–553 (2016). 131. Egawa N, Kitaoka S, Tsukita K et al. Drug screening for ALS using patient-specific induced pluripotent stem cells. Sci. Transl. Med. 4(145), 145ra104–145ra104 (2012). 132. Burkhardt MF, Martinez FJ, Wright S et al. A cellular model for sporadic ALS using patient-derived induced pluripotent stem cells. Mol. Cell. Neurosci. 56, 355–364 (2013). 133. Sareen D, O’Rourke JG, Meera P et al. Targeting RNA foci in iPSC-derived motor neurons from ALS patients with a C9ORF72 repeat expansion. Sci. Transl. Med. 5(208), 208ra149 (2013). 134. Wainger BJ, Kiskinis E, Mellin C et al. Intrinsic membrane hyperexcitability of amyotrophic lateral sclerosis patient-derived motor neurons. Cell Rep. 7(1), 1–11 (2014). 135. Kiskinis E, Kralj JM, Zou P et al. All-optical electrophysiology for high-throughput functional characterization of a human iPSC-derived motor neuron model of ALS. Stem Cell Rep. 10(6), 1991–2004 (2018). 136. Madill M, McDonagh K, Ma J et al. Amyotrophic lateral sclerosis patient iPSC-derived astrocytes impair autophagy via non-cell autonomous mechanisms. Mol. Brain. 10(1), 22 (2017). 137. Ferraiuolo L, Meyer K, Sherwood TW et al. Oligodendrocytes contribute to motor neuron death in ALS via SOD1-dependent mechanism. Proc. Natl Acad. Sci. USA 113(42), e6496–e6505 (2016). 138. Thompson AJ, Baranzini SE, Geurts J, Hemmer B, Ciccarelli O. Multiple sclerosis. Lancet 391(10130), 1622–1636 (2018). 139. Bechler ME, Swire M, Ffrench-Constant C. Intrinsic and adaptive myelination-a sequential mechanism for smart wiring in the brain. Dev. Neurobiol. 78(2), 68–79 (2018). 140. Lee S, Leach MK, Redmond SA et al. A culture system to study oligodendrocyte myelination processes using engineered nanofibers. Nat. Methods 9(9), 917–922 (2012). 141. Lariosa-Willingham KD, Rosler ES, Tung JS, Dugas JC, Collins TL, Leonoudakis D. Development of a central nervous system axonal myelination assay for high throughput screening. BMC Neurosci. 17(1), 16 (2016). future science group www.future-science.com 1321 Review Garcia-Leon, Vitorica & Gutierrez 142. Deshmukh VA, Tardif V, Lyssiotis CA et al. A regenerative approach to the treatment of multiple sclerosis. Nature 502(7471), 327–332 (2013). 143. Mei F, Fancy SPJ, Shen Y-AA et al. Micropillar arrays as a high-throughput screening platform for therapeutics in multiple sclerosis. Nat. Med. 20(8), 954–960 (2014). 144. Li Z, He Y, Fan S, Sun B. Clemastine rescues behavioral changes and enhances remyelination in the cuprizone mouse model of demyelination. Neurosci. Bull. 31(5), 617–625 (2015). 145. Cree BAC, Niu J, Hoi KK et al. Clemastine rescues myelination defects and promotes functional recovery in hypoxic brain injury. Brain 141(1), 85–98 (2018). 146. Green AJ, Gelfand JM, Cree BA et al. Clemastine fumarate as a remyelinating therapy for multiple sclerosis (ReBUILD): a randomised, controlled, double-blind, crossover trial. Lancet 390(10111), 2481–2489 (2017). 147. Kerman BE, Kim HJ, Padmanabhan K et al. In vitro myelin formation using embryonic stem cells. Development 142(12), 2213–2225 (2015). 148. Najm FJ, Madhavan M, Zaremba A et al. Drug-based modulation of endogenous stem cells promotes functional remyelination in vivo. Nature 522(7555), 216–220 (2015). 149. Hubler Z, Allimuthu D, Bederman I et al. Accumulation of 8,9-unsaturated sterols drives oligodendrocyte formation and remyelination. Nature 560(7718), 372–376 (2018). 150. Gumpel M, Lachapelle F, Gansmuller A, Baulac M, Baron van Evercooren A, Baumann N. Transplantation of human embryonic oligodendrocytes into shiverer brain. Ann. NY Acad. Sci. 495, 71–85 (1987). 151. Roy NS, Wang S, Harrison-Restelli C et al. Identification, isolation, and promoter-defined separation of mitotic oligodendrocyte progenitor cells from the adult human subcortical white matter. J. Neurosci. 19(22), 9986–9995 (1999). 152. Miron VE, Jung CG, Kim HJ, Kennedy TE, Soliven B, Antel JP. FTY720 modulates human oligodendrocyte progenitor process extension and survival. Ann. Neurol. 63(1), 61–71 (2008). 153. Ruffini F, Arbour N, Blain M, Olivier A, Antel JP. Distinctive properties of human adult brain-derived myelin progenitor cells. Am. J. Pathol. 165(6), 2167–2175 (2004). 154. Br¨ ustle O, Choudhary K, Karram K et al. Chimeric brains generated by intraventricular transplantation of fetal human brain cells into embryonic rats. Nat. Biotechnol. 16(11), 1040–1044 (1998). 155. Cummings BJ, Uchida N, Tamaki SJ et al. Human neural stem cells differentiate and promote locomotor recovery in spinal cord-injured mice. Proc. Natl. Acad. Sci. USA 102(39), 14069–14074 (2005). 156. Pluchino S, Gritti A, Blezer E et al. Human neural stem cells ameliorate autoimmune encephalomyelitis in non-human primates. Ann. Neurol. 66(3), 343–354 (2009). 157. Windrem MS, Schanz SJ, Guo M et al. Neonatal chimerization with human glial progenitor cells can both remyelinate and rescue the otherwise lethally hypomyelinated shiverer mouse. Cell Stem Cell. 2(6), 553–565 (2008). 1322 Future Med. Chem. (2019) 11(11) future science group