The application of induced pluripotent stem cell in neurological diseases

Zhichun Chen^1,2

1.     The Institute of Brain Science, FudanUniversity

2.     The State Key Laboratory of MedicalNeurobiology, Fudan University

Abstract

      Neurologicaldisease is a big challenge for modern medical field. Some of the neurologicaldiseases lack of early diagnostic biomarkers and disease-modifying therapies. Thisfrustrated situation is due to the poor understanding of disease pathogenesisand limited research approaches. Recent induced pluripotent stem cell (iPSC)technology has revolutionized our thinking patterns about the disease modelingand drug screening in neurological diseases. New pathogenic mechanisms havebeen dissected in Motor neuron disease, Parkinson’s disease, Alzheimer’s disease, and Huntington’s disease using iPSC. In addition, iPSC also provides highly efficientand reliable drug screening platform to assist developing new therapeutics for neurologicaldiseases. Furthermore, with the advantage of lack of immune rejection aftertransplantation, iPSC also seems to be a promising stem cell resource for cellplacement therapy. This review discusses the promises and challenges of iPSCapplication in future intervention of neurological diseases.

Keywords:neurological disease; inducedpluripotent stem cell; disease modeling; therapy; drug screening

1. Introduction

Neurological disease is characterized by the progress impairment ofneural cells involved in different anatomical and functional systems,accompanied by clinical abnormalities, including cognitive impairment, movementdeficiency, sensory impairment, emotional dysfunction, and other variousabnormalities[1]. Except stroke, neurodegenerativedisease is a neurological disease with significantly high disease burden[2]. According to the 2005 report from theWorld Health Organization (WHO), neurodegenerative diseases such as Alzheimer’s disease (AD), other dementias andParkinson’s disease (PD)made up more than 14% of the global burden of neurological illnesses[3]. Based on the data from Alzheimer’s Association, it is estimated to have5.4 million AD patients in US, more than 6 million patients in China, and 35.6million patients worldwide[4]. According to Alzheimer’s Disease International (2010), thetotal worldwide societal cost for dementia is nearly $604 billion in 2010,which is 1% of global gross domestic product (GDP), and is predicted to increase85% in 2030[5]. Amyotrophic lateral sclerosis (ALS) isa very severe motor neuron disease, upon diagnosed, the patients only havethree to five year to survive[6]. Both AD and ALS lack ofdisease-modifying therapies and current drugs can’t block the disease progression[6, 7]. Even though PD seems to have L-DOPA, dopaminereceptor agonists, deep brain stimulation for treatment, all of which can causea series of side effects, and after a period of application, the symptoms willcome back again[8-12]. As for Huntington’s disease(HD), a genetic neurologicaldisease, there is still no effective therapy for it now[8]. Thus developing effective approachesto treat these neurological diseases is a pressing concern in today’s world.

Generally speaking, the concrete pathogenesis of AD, PD, ALS, and HD isstill unknown. Though β-amyloid (Aβ) seems tobe the culprit of AD neurodegeneration based on the in vitro and in vivoexperiments demonstrating its neurotoxicity and induction of multiplepathogenic cascades, recent Aβ targeting drugs using γ-secretase inhibitors areall found to be ineffective to improve the outcomes of the AD patients, andeven the Aβ immunotherapy that cleared all of the extracellular Aβ depositioncan’t prevent disease progress, making AD a really enigmatic disorder[13-17].PD is caused by the selective loss of dopaminergic neurons in midbrain,resulting in the deficiency of dopamine transmission in striatum and subsequentmotor and non-motor symptoms[18]. However, although 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) or 6-hydroxydopamine have been used formodeling of PD, the pathophysiology of sporadic PD remains to be controversial[18-20].The most surprising thing is how motor neurons are selectively impaired insporadic ALS(SALS) in the condition that gene mutations that occurring infamily ALS(FALS) is rare in SALS[21]. Although the gene alterations responsible for HD had been identifiedfor over ten years, the particular cellular events leading to the death ofmedium spiny neurons in striatum are not clear[22]. Transgenic mice models can provide some invaluable information forunderstanding disease pathogenesis, however, in most proportion ofneurodegenerative diseases, the sporadic type did not contain the mutated geneor other genetic abnormalities. Furthermore, because of the individual andspecies difference in genetic background, the animal model can’t completelyreflect the specific pathogenic process of neurodegenerative diseases inindividual humans. Recent progress in stem cell technology, especially thereprogramming technology provides possibility to transform somatic cells intopluripotent stem cell or another type of cells. Induced pluripotent stemcell(iPSC) is developed by shinya Yamanaka and colleague in 2006, they use fourtranscript factors, Sox2, Oct4, Klf4, c-Myc, to transform the fibroblasts intopluropotent stem cells in mouse[23].This milestone study opened the world of reprogramming and make it possible todevelop cell model of neurological diseases. Due to the same genetic backgroundof iPSC to the subjects providing the somatic cell used for reprogramming, theiPSC will show the similar pathogenic process to those living in their bodiesunder similar pathogenic environment, which makes disease modeling using iPSC apromising technique for the deciphering of disease pathogenesis[24-26].

iPSC is also apromising platform for drug screening in specific patient, providing insightsfor the development of future personal medicine[27, 28].iPSC has the same genetic background to the patient who proffer the cellresource for reprogramming, which may be a perfect tool to screen potentialtherapeutic pharmacies, as well as to testify the safety and efficacy of thedrugs in neurological diseases[28]. In fact, some studies actually have used iPSC for the development ofnew drugs, like AD and ALS[29-31].

Finally, theiPSC seems to be also a perfect stem cell resource for transplantation in thetreatment of neurological diseases[32-34].As the iPSC contains same antigens that the patient has in his own body, afterthe transplantation of iPSC, it would not cause immune rejection and avoid theuse of immune suppressors[34-36].Using iPSC for transplantation treatment will not evoke ethical problems thatoccur in the condition of fetal stem cell or embryonic stem cell[37]. A number of studies had used iPSC for the treatment of someneurodegenerative diseases, including AD and PD[38-40].

In all, recentprogress in iPSC has revolutionized our understanding about neurological diseases.In this review, I will discuss the application of iPSC for the illumination ofdisease pathophysiology, drug screening, and transplantation therapy in fourneurological diseases.

2. Neurologicaldisease: disease burden and challenges

Alzheimer’s disease

AD is the most common type of neurodegenerative diseases[2]. Apart from 5% to 10% of AD withgenetic abnormality, about 90% of the AD is sporadic, lacking the mutations ofgenes related to family AD[41]. The major clinical manifestations ofAD is gradual cognitive impairment with age increasing[41]. As the disease progresses, thepatients may have personality or behavioral changes[42, 43]. In general, the clinical phenotype isthought to be related to the pathological loss of synapse and neurons, as well asthe subsequent brain atrophy[41]. Currently, the causes resulting inthe pathology are still unknown. In 1992, Hardy and Selkoe founded the core ofthe well-known amyloid cascade hypothesis[44], which argues that the excess Aβ peptideproduced in AD brain may induce multiple pathological alterations, includingmitochondrial dysfunction, oxidative stress, excitotoxicity and so on[45]. The accumulation of Aβ forms the amyloid plaques, which is one of the pathologicalha1lmarks of AD and can be detected by Pittsburgh Compound B（PiB） using positron emissionimaging (PET)[46, 47].Other toxic factors, especiallyhyperphosphorylated tau also plays an essential role in the pathogenesis of AD[48]. The hyperphosphorylated tau tends toaggregate and forms neurofibrillary tangles, which is also the pathologicalfeature of AD[49]. Though acetylcholinesteraseinhibitors and anti-excitotoxicity drugs have been used for the treatment of AD,there is no disease-modifying therapies currently[41, 50].

The most surprising thing in AD is that the most promising anti-Aβ therapy is demonstrated to be no efficacy to improve the outcomes ofAD in recent large-scale clinical trials[13, 14, 16, 51].This frustrated situation has raised new explorations on the basic initiatorsof pathophysiology in AD. Let’s think this complicated problem step by step,lots of evidence has demonstrated the neurotoxicity and synaptotoxicity of Aβin vitro and in vivo[52-54],thus Aβ surely plays important role in AD pathogenesis, considering the imagingdata that AD patients showed high PiB binding in their cortice[46, 55, 56].Now let’s think on the angle of one neuron, or astrocyte, for them, theextracellular Aβ is only one component of the external environments. That is tosay, even the extracellular Aβ is eliminated by immunotherapy, other toxicinsults, such as inflammatory factors, free radicals, activated astrocytes,reactive T cell, B cell, and microglia, all of which may continue to causeneurotoxicity and synapse impairment[57]. Besides, even the extracellular microenvironment is improved throughthe clearance of Aβ, the intracellular Aβ can continue to cause cellulardysfunction[54, 58].Furthermore, even though the Aβ has been cleared, the hyperphosphorylated tauis still there to exert its neurodegenerative effects[59]. Finally, when AD has been diagnosed and when AD patients showedclinical symptoms, the neurons that seems to be alive may have been irreversiblyimpaired, thus even the Aβ elimination may be ineffective for the delay of thedisease progression[60]. Glucose metabolism dysfunction is also a pathological feature of AD,especially insulin resistance and intracellular glucose metabolism maycontribute to the neurodegeneration, too[60]. All of these possible mechanisms may be involved in the futuredevelopment of therapeutics in AD.

Parkinson’s disease

PDis a chronic and progressive neurodegenerative disorder of the brain thataffects movement; emotional and cognitive deficits emerge in later stages[18, 61]. It usually develops in individualsover 50 (mean age is 60); Rare familial forms are known, but most cases aresporadic, too[62]. Environmentally induced forms havealso been described. PD is the most common motor neurodegenerative disorder andis second most common neurodegenerative disorder after AD[62]. Prevalence is 1% at age 60; 4-5% at85[63]. The “Cardinal” symptoms of PD include tremor, stiffness,bradykinesia, and postural instability[62]. Other symptoms may also occur, suchas slow speech, loss of expressiveness in face; diminished sense of smell,sleep disorders, cognitive deficits, depression, apathy, anxiety, dementia[62, 64]. PD is generally thought to be causedby the complete, or nearly complete loss of dopamine-producing neurons in thesubstantia nigra pars compacta[62]. The deposits of aggregated protein(Lewy bodies), containing denatured alpha-synuclein and other proteins, arealso a pathological hallmark of PD[62, 65]. There is a growing consensus that PDis heterogeneous, with subtypes distinguished by age of onset, clinicalsymptoms and progression rates[61, 62, 66]. About 15% of PD cases are inherited,and 17 PD susceptibility chromosomal loci (PARK1-17) and at least 10 specificPD susceptibility genes for familial PD have been identified[62, 67, 68]. The remaining 85% of PD cases are “idiopathic”, but PD susceptibility genes have alsobeen implicated in a small proportion of sporadic cases[67]. The environmental factors may includeherbicides and insecticides (e.g., paraquat, rotenone), manganese(e.g., maneb,ferro-manganese), metals(mercury, copper, lead), organic solvents(toluene,N-hexane, carbon disulfide, trichloroethylene), viral infections(post-encephalytic parkinsonism),physical trauma (brain)[69, 70]. Symptomatic relief of movementdeficits in PD (tremor, rigidity, akinesia) is realized by systemicadministration of L-DOPA, a dopamine precursor[18]. Dopamine receptor agonists are alsoused for treatment of PD, including bromocriptine, ropinirole, pramipexole,piribedil, apomorphine, lisuride, rotigotine[12]. Another type of therapy for PD is thestimulation of neurons in the STN or GPi, namely deep brain stimulation (DBS),which can suppress involuntary movements (e.g., chorea, dystonia) that arise asside-effects of L-DOPA therapy[11, 71].

The five major genes implicated in familial PD is SNCA(alpha-synuclein) and LRRK2 (Leucine-rich repeat kinase 2) in late onset,autosomal dominant inheritance PD, and PRKN (parkin, PARK2: E3 ubiquitinligase), PINK1 (PARK6; PTEN-induced putative kinase 2), DJ-1 (PARK7; molecularchaperone/anti-oxidant) in early onset, autosomal recessive inheritance PD[62, 67, 68]. How these gene abnormalities inducephenotype of PD has been investigated in a spate of studies. The mechanismsimplicated in PD can be classified into cell-autonomous mechanism and non-cellautonomous mechanism. The cell-autonomous mechanism includes mitochondrialdysfunction, protein aggregation, increased Ca2+ influx, oxidative damage dueto ROS production, and autophagy or mitophagy[72, 73]. Excitotoxicity, inflammation,including production of apoptosis-promoting cytokines by microglia are thecommon non-autonomous mechanisms[73, 74]. Though L-DOPA, dopamine receptoragonists, and DBS may provide effective therapeutic effects to PD patients, howthe dopaminergic neurons are lost in PD is still not clear, and why othersystems also participate in PD progress is also an open question. Furthermore,for a great number of PD patients, current therapies are all temporallyeffective, five years or ten years later, the disease may continue to exacerbate,which may require more researches on the development of PD[62].

Amyotrophiclateral sclerosis

ALS is a progressive, fatal motor neuron disease (MND) caused by thedegeneration of motor neurons in the motor cortex, brain stem and spinal cord[21]. The symptoms of ALS include muscleweakness and atrophy throughout the body; muscle fasciculation, cramping andstiffness; difficulty in speaking (dysarthria) and swallowing (dysphagia)[21]. A small percentage (~5%) of ALSpatients develop symptoms of frontotemporal lobar degeneration  (ALS-FTLD)[75]. It commonly strikes between the agesof 40 and 60. SALS occurs in ~90 - 95% of cases. FALS accounts for theremaining 5-10%. About 1/5 of these are caused by a dominant mutation in thesuperoxide dismutase gene (SOD1) on chromosome 21[21]. Mutations in TARDBP, encoding TAR DNAbinding protein 43 (TDP43), and FUS, encoding fused in sarcoma/translated inliposarcoma (FUS/TLS) protein, have been linked to rare forms of FALS or SALS[76, 77]. Recent discovery of C9orf72, abnormalexpansions of GGGGCC-repeat in intron 1, is the most common gene implicated in ALS currently[78, 79]. Riluzole is the only drug forclinical treatment of ALS, however, it only can delay the disease progressionfor 3 to 4 months[6, 80]. There is still no disease-modifyingtherapies for ALS.

The role of SOD1 in ALS has been studied for two decades, and the SOD1transgenic animal is the most common model for ALS research[81, 82]. However, SOD1 represents only a smallpercentage of FALS[21], and the role of TDP43, C9orf72, andFUS/TLS may need to be studied to decipher the pathogenesis of ALS. Recentgreat progress in ALS is the emerging correlation between ALS and FTD,represented by the genes occurring in both ALS and FTD include C9orf72, VCP,SQSTM1 (p62), OPTN, UBQLN(ubiquitin-like protein, ubiquitin 2)[83, 84]. How these genes can both participatein the deficits of movement and impairment of cognition or personality isreally an interesting question to explore in the future.

Huntington’s disease

HD is a progressive neurodegenerative disease that causes movementdisorders, cognitive difficulties, and psychiatric disturbances[22, 85]. Onset of clinical symptoms in HD isusually in middle age (35-44). The prevalence of HD for European andEuropean-descent is 5-10 per 100,000, but for Asian, African and nativeAmerican populations, the prevalence is lower with 0.5 per 100,000. The patientsurvival is usually 15-25 years after diagnosis[86]. However, juvenile form is also knownin Westphal variant. HD is inherited as a dominant gene and 50% of offspringare affected[86]. The clinical phenotype is generallycaused by the mutation of huntingtin (HTT) gene located in chromosome 4[22, 87]. HTT is expressed ubiquitously, butmutant forms containing 40 or more CAG triplets in exon 1 (encoding glutamine)are toxic to neurons, especially medium spiny projection neurons in thestriatum (caudate nucleus and putamen)[22]. Expansion of repeats occurs morefrequently during spermatogenesis compared to oogenesis, so that pathogenicalleles are often inherited from fathers[22, 88]. Also, repeats of 40 or more areunstable and tend to increase from generation to generation. This phenomenonexplains genetic “anticipation”: the earlier onset of HD in offspringcompared to the parent[89]. The possible mechanisms to explaingenetic dominance of mutant HTT(mHTT) include “haploinsufficiency”, dominant-negative effects, and toxic “gain-of-function”[85, 90]. The mechanism for the selectivedegeneration of caudate and putamen neurons may be due to the formation ofmHTT/Rhes complex by mHTT and GTP-binding protein Rhes, which causes theSUMOylated mHTT, resulting in the mHTT disaggregation and subsequent cell death[91]. The treatment of HD is primarily thereduction of symptoms. The movement disorder (chorea) can be treated bytetrabenazine (a VMAT-inhibitor), haloperidol (antipsychotic; D2 receptorantagonist), clonazepam (anticonvulsant; modulates GABAA receptor)[22, 86]. The psychiatric symptom can betreated by haloperidol (antipsychotic), fluoxetine, sertraline (SSRIs), andnortriptyline (SNRI)[22, 86].

Though the genetic origin of HD has been found, however, the therapytargeting HTT aggregates is still being developed. Besides, how the diseaseprogress and the respective role of different cellular events occurring in HDare also fields to be explored.

3. Inducedpluripotent stem cell: principles and prospects

Generation of inducedpluripotent stem cell

The milestone investigation in iPSC is established in 2006 and 2007, byTakahashi and Yamanaka, which was done by introducing four reprogramming genes,Oct4, Sox2, Klf4, and c-MYC, into differentiated somatic cells[23, 92]. This approach made it possible (withsome exceptions) to establish iPS cells from the somatic cells of any individual,regardless of race, genetic background, or state of health (i.e., whetherafflicted with a disease). Moreover, the development of in vitrodifferentiation protocols for embryonic stem cells toward each type of thecells in the body provide the technique for iPSC differentiation into othercells, such as heptocyte, cardiocyte, neuron and so on[92]. The advantages of the iPSC is thatthe derived cell had the same genetic background to the patient who derives it,which is a perfect in vitro duplication of the patient.

The first generation of human iPSCs (hiPSC) was established usingretroviruses and lentiviruses to introduce a cocktail of transcription factorsfor reprogramming[92, 93]. Both retroviral and lentiviralvectors can assist inserting the transgene into the host cell genome andectopic expression of the transcription factors needs the reactivation ofendogenous genes that are necessary for pluripotency. There are severaldisadvantages to use retroviral and lentiviral gene delivery systems for hiPSC generation[94, 95]. The multiple proviral integration,incomplete silencing or reactivation of the transgenes all may increase thepossibility of tumorigenesis, particularly when c-Myc, a proto-oncogene, isused in the reprogramming cocktail[96-98]. Furthermore, there is also aninherent risk of malignant transformation because of the use of retroviral orlentiviral vectors for gene transfer[99]. Currently, reprogramming strategieshave emphasized the safety of reprogramming and the expectation to producehiPSCs with minimal disturbance to the host cell genome. In order to reduce thetumorgenic potential of the conventional Yamanaka factors, hiPSCs have beenproduced only by three of the four factors, lacking the c-Myc[100, 101]. HiPSCs have also been derived bysubstituting Klf4 and c-Myc with Lin28 and NANOG transgenes in the four-factorcocktail[93]. Additionally, hiPSCs have also beengenerated by overexpressing Sox2 and Oct4 by adding valproic acid, a histonedeacetylase inhibitor, which has a similar efficiency to the three-factorcocktail[102]. Either nonintegrating viral vectors(e.g., adenoviruses and Sendai virus) or physical methods of gene transfer(e.g., electroporation of plasmids that remain episomal) have been developed asmethods to derive hiPSCs[103, 104]. Though temporary expression ofreprogramming transcription factors through nonintegrating methods cansuccessfully generate bona fide hiPSCs, its efficiency is extremely low[105]. It has been reported if using thehuman fetal neural progenitor cells as the starting material can enhance theefficiency of episomal-based reprogramming[106]. If directly using the neural stemcells as starting material, which have endogenous expression of Sox2 and c-Myc,it may only require Oct4 and NANOG for reprogramming[106]. A single nonviral vector containingall four Yamanaka factors has been utilized to derive hiPSC by combining withpiggyBac transposons that can expel the plasmid from the cell genome uponreprogramming has started[107]. Delivery of membrane-permeable taggedrecombinant proteins and synthetic mRNAs of reprogramming factors into the cellis another successful way for reprogramming, but the efficiency is low[108]. Recently, the administration of smallmolecules for iPSC establishment instead of the introduction of transgenes hasbeen developed. This method increased the efficiency by 200 folds[109].

Characterization ofinduced pluripotent stem cell

The validation of complete reprogramming and conformation ofdevelopmental functionality of hiPSCs is a long, arduous and unavoidableprocess[110]. The existence of a high percentage ofincompletely reprogrammed cells, which may form tumor,  makes the selection and identification of afully reprogrammed hiPSC colony very necessary and important[110]. Colonies are initially selected basedon whether the cell line has a similar morphology to hESCs. The retroviralsilencing and cytogenetic analysis are confirmed by using the standard assaysof alkaline phosphatase staining, detecting pluripotency markers, measuring DNAmethylation status of promoters of pluripotent genes[92, 110]. To demonstrate its pluripotency, thecells are induced to differentiate in vitro into various mature cell typesrepresenting the three germ layers, typically using an embryoid bodyintermediate[111, 112]. Teratoma formation is also used forthe confirmation of pluripotency, by injecting the stem cell intoimmunocompromised mice to test whether it can form the tumors that contain celltypes originating from the mesoderm, ectoderm and endoderm[113]. In mouse, the generation ofgermline-competent chimeric mice is often used, or more strictly, a viable miceis required[114, 115]. However, this can’t be done in hiPSC because of thearising of ethical issues.

Neural differentiationof induced pluripotent stem cell

iPSC can be efficiently differentiated into a variety of neuronal ornonneuronal fates by using various differentiation protocols. These protocolsoften make full use  of existing in vivopathways that drive mammalian CNS embryonic development. Some protocols forneural stem cell differentiation are based on concurrent inhibition of twoparallel SMAD/ transforming growth factor-β (TGFβ) superfamilysignaling pathways—mediated bybone morphogenic proteins (BMP) and Activin/Nodal/TGFβ[116]. Both pathways are responsible for the induction of nonneuronal fates,such as epidermis or mesoderm during CNS development. Then by using anothersignaling pathways involved in the specific differentiation of neural cells maysupport the differentiation of diverse neuronal types, like glutamatergicneurons in basal forebrain [117].For many neurological diseases, one specific neural subtype is more susceptibleto be affected, such as cholinergic neurons in AD[118], motor neurons in ALS[21], dopaminergic neurons in PD[62]and medium spiny neurons in HD[86]. Thus many studies have achieved to produce disease-specific neurontypes through different protocols. For example, dopaminergic neurondifferentiation from iPSCs can range from 21 days to more than 2 months inculture[119]; functional motor neuron differentiation requires 8-10 weeks[120]and medium spiny neuron differentiation takes more than 60 days[121].The capacity to generate specific types of neurons from patient-derived hiPSCsholds much promise for disease modeling and the understanding of the molecularand cellular mechanisms underlying neuropathology.

4. Inducedpluripotent stem cell for modeling and drug development of neurodegenerativedisease

Alzheimer’s disease

As what Ianalyzed above, intracellular Aβ, hyperphosphorylated tau, glucose metabolismdysfunction may be involved in the AD pathogenesis. To dissect their roles inAD, iPSC may be a preferential platform to reveal the mechanisms of differentpathologies. Recent iPSC study in AD have observed the Aβ accumulation in thecells derived from patients with family type of AD[122, 123].Presenilin 1(PSEN1) and presenilin 2 are the components of intramembraneprotease complex of γ-secretase, which participates in the processing ofamyloid precursor protein(APP). β-secretase is also involved in the Aβproduction. Both γ-secretase inhibitors and β-secretase inhibitors seem to havesuppressive effects on the Aβ production in the published iPSC studies of AD[122, 123].One study also showed the production of Aβ oligomers and induction ofendoplasmic reticulum stress and oxidative stress, and docosahexaenoic acidtreatment can attenuate the effects of stress response in one of the patients,but not in another patient[124].Zhang et al(2014) demonstrated how a three dimensional (3D) culture of stemcell derived neurons can induce in vivo like responses related to AD, but notrecapitulated with conventional 2D cultures, which indicates that extracellularmicroenvironment is also very important for the pathogenesis of AD[125]. Sproul et al (2014) also studied the molecular profiling of PSEN1 FADusing iPSC, they found increased generation of Aβ42/40 and characterized novelexpression changes in the PSEN1 iPSC model[126].Muratoreet al (2014) produced iPSC from patients harboring the London FAD APP mutation(V717I). They found that β-secretase cleavage of APP is elevated resulting ingeneration of increased levels of both APPsβ and Aβ and this mutation also altersthe initial cleavage site of γ-secretase, leading to an increased generation ofboth Aβ42 and Aβ38[127]. They also observed that treatment with Aβ-specific antibodies earlyin culture rescued the phenotype of increased total tau levels. The studies onthe drug screening and iPSC transplantation or iPSC derived celltransplantation in AD are rare now, highlighting the necessity of encouragementfor iPSC investigations in AD. The figure 1. shows that differentiation of iPSCinto various types of neurons for neurological disease utilization.

Figure1. The differentiation of iPSC into various disease-specificneurons.

Parkinson’s disease

Nguyenet al (2011) demonstrated that iPSC carrying the G2019S mutation in the LRRK2gene showed increased expression of key oxidative stress-response genes andα-synuclein protein[128]. Ascompared with control DA neurons, G2019S-iPSC is more sensitive to caspase-3activation and cell death induced by stress agents, including hydrogenperoxide, MG-132, and 6-hydroxydopamine[128].Soldner et al (2011) genetically corrected disease-causing point mutations inpatient-derived hiPSCs by combining zinc finger nuclease (ZFN)-mediated genomeediting and iPSC technology, which may be used for the further confirmation ofrelationship between mutated gene and phenotype of iPSC[129].Similar technique is also used by Sanders et al, who realized ZFN-mediatedrepair of the LRRK2 G2019S mutation in iPSCs. Byers et al(2011) found thatiPSC-derived midbrain dopaminergic neurons from a patient with a triplicationin the α-synuclein gene (SNCA) exhibited disease-related phenotypes in culture,characterized by accumulation of α-synuclein, oxidative stress, as well assensitivity to peroxide induced oxidative stress[130]. Sánchez-Danés et al (2011) reported that dopaminergicneurons differentiated from either idiopathic PD- or LRRK2-PD-iPSC showedmorphological alterations, including reduced numbers of neurites and neuritearborization, as well as accumulation of autophagic vacuoles[131]. Cooperet al (2012) analyzed neural cells generated from iPSC derived from PD patientsand presymptomatic individuals carrying mutations in the PINK1 (PTEN-inducedputative kinase 1) and LRRK2, revealing the importance of oxidative stress andmitochondrial dysfunction in PD pathogenesis[132]. Studyalso found that PARK2 iPSC-derived neurons showed increased oxidative stressand enhanced activity of the nuclear factor erythroid 2-related factor 2 (Nrf2)pathway, and exhibited abnormal mitochondrial morphology and impairedmitochondrial homeostasis, but not Lewy body (LB) formation with anaccumulation of α-synuclein[133]. Ryan etal (2013) reported that redox reaction inhibited the MEF2C-PGC1α transcriptionalnetwork, contributing to mitochondrial dysfunction and apoptotic cell death inA53T α-synuclein (α-syn) mutant cells derived from iPSC, providing mechanisticinsight into gene-environmental interaction (GxE) in the pathogenesis of PD[134]. Milleret al (2013) expressed progerin in iPSC-derived fibroblasts and neurons tomodel late-onset PD and their results showed that dopamine neurons presentedpronounced dendrite degeneration, gradual loss of tyrosine hydroxylase (TH)expression, and enlarged mitochondria or LB-precursor inclusions[135]. Schöndorfet al (2014) generated dopaminergic neurons from iPSC derived from patient withmutations in the acid β-glucocerebrosidase (GBA1) gene, which is a risk factorof PD[136]. Theyobserved that mutated neuron showed a dysregulation of calcium homeostasis andincreased vulnerability to stress responses related to elevation of cytosoliccalcium, which can be rescued by the mutated gene correction[136].

    Swistowskiet al (2010) reported the first study on the efficient generation of functionaldopaminergic neurons under defined conditions using iPSC[137]. They also demonstrated that iPSC-derived dopaminergic neurons werefunctional because they survived and improved behavioral deficits in6-hydroxydopamine-leasioned rats after transplantation. Kikuchiet al (2010) found that midbrain dopaminergic neurons produced by a feeder-freeneural differentiation method from human iPSCs can be grafted into the brainsof a monkey PD model and NOD-SCID mice[138].Another group of researchers also demonstrated that midbrain-like DA neuronsgenerated from  iPSCs of cynomolgusmacaque integrated into the striatum of a rodent model of PD and promotedbehavioral recovery[139]. Similar results are found that primate iPSC-derived neural cellssurvived in the striatum of one primate without any immunosuppression[35]. Byusing different methods, studies also showed that hESC/iPSC geneticallyengineered with lentiviral vectors driving controlled expression of LMX1A is anefficient way to produce enriched populations of human A9-subtype DA neurons inventral brain, which may implicate for the drug screening and resource for transplantation[140]. Mar etal (2011) found that small molecules greatly promoted conversion of human iPSCinto the neuronal lineage, which may potentially provide more simple approachfor iPSC establishment[141].

Amyotrophiclateral sclerosis

Dimos et al (2008) firstly demonstrated that iPSC from ALS patients canbe differentiated into motor neurons[142]. Mitne-Netoet al(2011) reported the second iPSC study in ALS, and they found reducedlevels of vamp-associated protein B/C (VAPB) protein in ALS8-derived motor neuronsdifferentiated from iPSC derived from FALS caused by mutations in the VAPB gene[143]. In 2012, Egawaet al. published the study on the drug screening for ALS using iPSC. Theygenerated motor neurons from iPSC from FALS patients, who carry mutations inTDP-43. Then they analyzed four chemical compounds and found that anacardicacid, a histone acetyltransferase inhibitor could rescue the abnormal ALS motorneuron phenotype[31]. In 2013, Popescuet al. found that iPSC-derived neural progenitors efficiently engrafted in theadult spinal cord and survived at high numbers in wild-type and transgenic ratscarrying a human mutated SOD1(G93A) gene, which may highlighting the promise ofiPSC for ALS cell therapy[144]. Another study using iPSC technologyin ALS found that motor neurons derived from three sALS patients exhibited denovo TDP-43 aggregation and that the aggregates are also confirmed inpostmortem tissue from one of the same patients from which the iPSC werederived. Nizzardo et al (2014) reported that minimally invasive injections ofiPSC-derived neural stem cell improved neuromuscular function and motor unitpathology and significantly increased life span of ALS mice[145]. Sareen et al (2013) repoted thatC9-ALS motor neurons derived from iPSC from ALS patients carrying the C9ORF72repeat expansion exhibited altered expression of genes involved in membraneexcitability including DPP6, and displayed less spikes upon depolarizationcompared to control motor neurons, which can be rescued by ASO targeting theC9ORF72 transcript[146]. Donnellyet al (2014) found that iPSC-differentiated neurons from C9ORF72 ALS patientsshowed disease-specific intranuclear GGGGCCexp RNA foci, dysregulated geneexpression, sequestration of GGGGCCexp RNA binding protein ADARB2, andsusceptibility to excitotoxicity, which can be reduced with antisenseoligonucleotide (ASO) targeted to C9ORF72 transcript or repeat expansion[147]. Chen et al(2014) revealed that spinalmotor neurons from patients with mutation in the SOD1 can exhibit neurofilament(NF) aggregation if glia were not present, which are associated with decreasedstability of NF-L mRNA and binding of its 3' UTR by mutant SOD1[82]. In addition, conditional expressionof NF-L can rescue the NF subunit proportion, diminishing NF aggregation andneurite degeneration.

Huntington’s disease

Juopperiet al (2012) condcuted the first iPSC study in HD, and the investigators  produced iPSCs from a father with adult onsetHD with 50 CAG repeats (F-HD-iPSC) and his daughter with juvenile HD with 109CAG repeats (D-HD-iPSC)[148]. Interestingly, they found thatneurons derived from both lines of iPSCs can be functional normal. However,when they differentiate HD-iPSC into an astrocytic lineage, they found thepresence of cytoplasmic, electron clear vacuoles in astrocytes from both linesof iPSCs. Thus they conclude that vacuolation may be a phenotype associatedwith HD. Another study in HD iPSC revealed that proteasome inhibiton may be apathological alteration in HD. Chae etal (2012) studied the proteomic profiles of HD-iPSCs and found that oxidativestress-related proteins, such as SOD1 and Prx (peroxiredoxin) families wereseriously affected in HD-iPSCs[149]. An etal(2012)found that iPSC derived from HD can be corrected by the replacement ofthe expanded CAG repeat with a normal repeat using homologous recombination,which repaired pathogenic HD signaling pathways and got rid of diseasephenotypes such as sensitivity to cell death and impaired mitochondrialbioenergetics in neural stem cells[150]. Another study also reporteddifferentiated HD neural cells from iPSC showed disease-associated changes inelectrophysiology, metabolism, cell adhesion, and ultimately cell death forlines with both medium and longer CAG repeat expansions[151]. The question is why the initial studyof iPSC in HD did not find the similar pathological alterations in HD neurons.The possibility is that the patients studied in first iPSC investigation in HDmay not show complete penetrance, and the genetic background may be resistantto the pathology. Fink etal (2014) demonstrated that intrastriatal transplantation ofadenovirus-generated iPSC in the brain of HD rats induced by 3-nitropropionicacid (3-NP) showed preserved motor function[152]. Jeon etal (2014) studied the therapeutic effects of HD patient-derived iPSC carrying72 CAG repeats in the YAC128 transgenic mice, a commonly used HD mouse modelcarrying 128 CAG repeats[153]. After the transplantation, theyobserved a significant behavioral improvement in the grafted mice, and after 12weeks transplantation, they did not find any EM48-positive protein aggregatesthat occurred in HD, which may suggest that HD pathology can’t be transmitted to the graft after 12week-transplantation[153].

5. Challenges andprospects on the application of induced pluripotent stem cell in neurologicaldisease

Diseasemodeling

The biggest advantage for iPSC in disease modeling is that the iPSCcarries the same genetic information to the patient who derives it. However, wehave to note that the iPSC may not contain the epigenetic modifications thataccumulated in the patients who derive it. Considering the environment has animportant role in the progress of sporadic disease, the iPSC possibly lost someimportant information that only can be got through long-term living experience.As iPSC is only a cell model, it is surely not a perfect prototype to recapitulatethe true conditions of the patients with the disease. Especially on the studyof non-cell autonomous mechanism of the disease, only by creating a similarmicroenvironment for the neural cells from iPSC, can we understand whathappened in the in vivo situation. As far as I am concerned, two lines ofexperiments can be performed to use iPSC as a tool for the study of diseasepathogenesis. The first line of the experiment focuses on the family type ofneurological disease with confirmed gene abnormality or gene polymorphism. Inthis case, the iPSC not only can help us to decipher the pathophysiologyinduced by the gene alteration, but help us know more about the pathologicalprocess occurring in sporadic disease. The other line of the experiment can bedesigned if there is suspected disease-related factor but previous studies didnot reveal the mechanism clearly. At the same time, iPSC also provide the cellmodel to testify the previous established conclusions about the disease.

Drugscreening

As I see, drug screening is the most promising application for iPSC.However, the problem is that in vitro condition can’t simulate the real states of drugabsorption, distribution, metabolism, and excretion, which may causemisunderstanding about the efficacy of drug in vivo. Especially for thepro-drugs, they may not directly affect the targeted cells in the body, buttheir product after metabolism is the one to interact with the cell and exertits therapeutic effects. Besides, the iPSC also can’t represent the true feelings of thepatient who take it, some patients may have allergy to the drug, or under someconditions, the drug may be not so tasty or with bad smelling, all of which can’t be found by the cell model. For somediseases, the iPSC is not suitable for its modeling with the reason that thepathological alteration occurred in the cell may be not the causal mechanism toinduce clinical phenotype, which may lead to the no efficacy of the drug in vivo.

Celltransplantation

The emerging of iPSC is a rescue for the field of stem celltransplantation. Previous studies on stem cell transplantation in diseasetherapy are confronted with at least four problems. The first one is that theembryonic stem cell or fetal stem cell may cause immune rejection aftertransplantation, which require the patients who grafted with the stem cell totake the immune-suppressors. This will lead to the low immunity of the patientto fight against with the microbes and tumors. The second one is that traditionalstem cell resource is unavoidably evoking ethical issues, even there is sometime when the stem cell research is not permitted. The third one is that eventhe stem cell is transplanted into the brain, they may not survive or notdifferentiate into the cell types that needed for disease rehabilitation. Thereis the so-called “seedand soil” hypothesis,even the seed, the stem cell, is functionally normal, after thetransplantation, the soil, namely the local microenvironment, may not be soappropriate for the stem cell to live. Finally, whether the transplanted cellcan restore the function of the patients is not clear. Especially for centralnervous system, even the cell can survive, if they did not repair the circuitsthat are specifically impaired in the disease, the function of the subjects maybe not completely recovered. However, we still believe that iPSC is the mostpromising cell line for future cell therapy, studies should be performed topropel its application in clinical trails for neurological disease treatment.

6. Concluding remarks

Neurologicaldisease is the most complicated and serious disease in the world. Its highprevalence, severe disease phenotype, and lack of early diagnosis and diseasemodifying therapies make it become one of the biggest challenges in thedevelopment of future society. iPSC, upon its invention, has become the hotspot among basic researchers and clinical practitioners who are devoted toutilize it for disease modeling, drug screening, and stem cell transplantation.Though problems are still in the road to make full use of iPSC technology,recent studies of iPSC on neurodegenerative diseases seem to highlight theunique value of iPSC in neuroscience research.