Induced pluripotent stem cells (iPSC) for modeling of Alzheimer’s disease

Publiceret Januar 2015

Stem cell-derived neurons can create “dementia” in a Petri dish.

Fibroblasts from patients with Alzheimer’s disease can be reprogrammed into induced pluripotent stem cells (iPSC), which are capable of differentiating into all cell types of the mammalian body. Hence, they can be differentiated into neurons, which amazingly express phenotypic characteristics of Alzheimer’s disease in the Petri dish. This allows for researchers to get close to the molecular disease mechanisms, which would normally be hidden deep inside the brain. The diseased neurons can also be utilized for drug discovery in order to find compounds that alleviate the impact of the disease processes. Furthermore, molecular tools allow for potential correction of mutations that induce the disease in familial cases. This points towards future cell-based therapy; a potential that will be realized for Parkinson’s disease already this year.

Alzheimer’s disease – a growing threat to life quality

Over the past decades, improvements in medicine has led to reducing early death caused by infectious diseases, cardiovascular diseases and even cancer. These achievements have resulted in an increase in the average lifespan by 23.7 years for men and 25.4 years for women over the past 100 years. This extended lifespan, however, poses a challenge related to the exponentially increasing risk to develop neurodegenerative diseases with advancing age. It is estimated that about 90.000 people in Denmark are affected by dementia, with Alzheimer’s disease (AD) being the most common form, and this figure is expected to rise to 160.000 by 2040. The prevalence of dementia is about 5% within the age of 65 to 74, but almost 50% for those above the age of 85. From these facts it is obvious that the ageing brain is more susceptible to neuronal degeneration, and that the risk of developing AD as well as other dementias increases steeply.

AD is a progressing and devastating brain disease for which no cure is available yet. AD can be divided into familial and sporadic forms. The familial forms, which account for only a few percentages of the cases, are caused by mutations in well-characterized genes, including the amyloid precursor gene (APP) and presenilin 1 and 2 genes (PSEN1, PSEN2). APP is an integral membrane protein, which probably acts in synaptic formation and repair. APP is cleaved by alpha-, beta- and gamma-secretases. PSEN1 and 2 are subunits of the gamma-secretase. Whereas APP is normally cleaved into Abeta 40 fragments, mutations in APP, PSEN1 or 2 may cause abnormal cleavage of APP in toxic Abeta 42 fragments, which results in formation of extracellular plaques in the brain typical for AD. In addition to the Abeta plaques, AD patients also display intercellular tangles of hyperphosphorylated Tau protein; a protein which is normally involved in the stabilization of the microtubules of the neurons. The interrelationship and causality of the Abeta and Tau aspects of the disease are still not clearly understood.

The familial AD forms are characterized by a younger age of onset, often between the ages of 40 to 50, whilst sporadic patients usually develop symptoms after the age of 65. So far the precise background for the sporadic cases is enigmatic, but certain gene variants, for e.g. apolipoprotein E4, have been identified as risk factors for developing AD, as well as life style and cardiovascular factors. So far, only symptomatic treatments of AD, developed 15-20 years ago, are available. In spite of great investments in potential disease modifying treatments, none of those has succeeded in demonstrating significant clinical efficacy. Hence, novel approaches for understanding of the molecular pathophysiology and for testing potential treatment strategies are urgently needed.

Induced pluripotent stem cells (iPSCs) – resetting the biological clock

Due to the inaccessibility of the brain, most pathophysiological knowledge on AD is accumulated via postmortem brain analyses and rodent models of the disease carrying human familial AD mutations. Post mortem brain analyses only provide a snapshot of the very late stage of AD. Transgenic rodent models, on the other hand, have not proven to be very successful in recapitulating human disease pathophysiology. Hence, we are in urgent need of more robust human disease models, which also recapitulate early AD disease phenotypes. After the groundbreaking discovery of induced pluripotent stem cells (iPSC) by Shinya Yamanaka in 2006 (Takahashi and Yamanaka, 2006), it became possible to take an adult cell, turn back its biological clock and revert it into a pluripotent cell resembling an embryonic stem cell. An iPSC has the ability to differentiate into any desired mature cell type, including specific neurons of the brain. This great discovery opened up the opportunity to develop in vitro models of patients with neurodegenerative disorders to study the disease pathology close up in the Petri dish with the obvious perspective for personalized cell-based therapy.

Shinya Yamanaka and John B. Gurdon were jointly awarded The Nobel Prize in Physiology or Medicine 2012 for their discovery that mature cells can be reprogrammed to become pluripotent.

Reprogramming to induced pluripotent stem cells (iPSCs) – think globally, act locally

Over the past years, a pipeline has been developed for utilization of the iPSC-based modeling of AD in Denmark, through projects supported by both the National Advanced Technology Foundation and the EU. Over the coming 6-year period, these activities will be boosted significantly though the stem cell center “BrainStem” supported by Innovation Fund Denmark and including partners at University of Copenhagen, Copenhagen University Hospital, University of Lund, University of Southern Denmark, Aarhus University, Bioneer A/S, Lundbeck A/S and the French company ICDD. Our activities have so far been focused on familial forms of AD, frontotemporal dementia and spinocerebellar ataxia, and well-characterized patients have been recruited after informed consent from the Danish Dementia Research Center (DDRC). Skin biopsies have been taken from the patients and the fibroblasts obtained from these samples have been submitted for reprogramming into iPSCs (Figure 1).

Figure 1: Schematic overview of potentials of induced pluripotent stem cells (iPSC) in relation to neurodegenerative disorders. Skin fibroblasts ate obtained from a Alzheimer's or Parkinson's disease patients, reprogrammed into iPSC and, in case of familial background, the mutations is corrected. Neurons are differentiated from diseased and healthy iPSC and compared for studies of molecular pathophysiology. The diseased neurons can also be utilized in drug development. Furthermore, healthy iPSC can be differentiated into neural progenitor cells as e.g. dopaminergic progenitors, which can be used for transplantation in the case of e.g. Parkinson's disease.

In Shinya Yamanaka’s groundbreaking experiments, retroviruses were used as vectors of the four reprogramming genes Oct4, Sox2, Klf4 and c-Myc, which are transcription factors required for cell pluripotency and self-renewal. When Yamanaka transduced mouse fibroblasts with the virus, the genes were integrated into the genome, and upon expression the fibroblasts were reprogrammed into iPSC. Hence, the fibroblast identity was lost and the biological clock reset to zero. However, integration of transgenes, of which some are oncogenic, is a considerable risk factor if cells are to be used for cell-based therapy. Hence, several subsequent reprogramming strategies have been developed with an increasing focus on avoiding genomic integration of transgenes, such as Sendaivirus, plasmids, RNA or proteins.

In our research, we have refined an episomal plasmid-based reprogramming technology (Rasmussen et al., 2014) in an attempt to lower the risk of transgene integration. We use three plasmids carrying the human sequences of OCT4, SOX2, KLF4, LIN28, L-MYC as well as a small hairpin RNA against p53 (shp53). All of these genes have previously been described to be essential for reprogramming and maintenance of the pluripotent ground state in iPSC. P53 is a tumor suppressor gene, which induces cell-cycle arrest or apoptosis in response to DNA damage. However, it has also been shown to exert an inhibitory role on reprogramming. The addition of the shp53 suppresses p53 expression and ensures that p53 cannot execute its inhibitory role, which ultimately results in proliferation and support of pluripotency. We have demonstrated that these effects occur without accumulation of chromosome abnormalities in the resulting iPSCs. In general, the plasmids are not integrated into the genome and leave the resultant iPSCs without transgenes.


Figure 2. Colony of induced pluripotent stem cells (iPSC)
after immunostaining for Nanog (pluripotency transcription
factor) and Occludin (cell adhesion molecule).
Counterstained for DNA. Photo: Miya Kudo Høffding

We have investigated the phase of reprogramming of the fibroblasts into the iPSCs carefully and described in detail how the individual cells from being fibroblasts of mesenchymal phenotype start to proliferate and undergo a mesenchymal-to-epithelial transition to form an epithelial colony consisting of pluripotent stem cells closely adhering to each other by intercellular junctions (Figure 2) (Hoeffding et al., 2014). During this process, mesenchymal genes are downregulated and epithelial and pluripotency-related genes are upregulated. After reprogramming, individual iPSC colonies emerge amongst the non-reprogrammed patient cells, and are subject to expansion. We routinely expand a minimum of six individual clones per patient and verify their pluripotency status through assessment of the expression of pluripotency genes, as well as confirmation of the absence of the episomal plasmids in the genome combined with a normal karyotype (Figure 3).

Figure 3. Colony of induced pluripotent stem cells
(iPSC) as visualized by scanning electron microscopy.
The iPSC colony is surrounded by fibroblasts that did
not undergo reprogramming. Note the close
apposition of the iPSC epithelium. Photo: Miya Kudo

In the case of familial AD, we aim at generating genetically edited iPSC lines where the disease-causing mutation is corrected. These so-called isogenic iPSC lines are optimal controls for in vitro cell modeling as they represent the same genetic background except for the mutation. In general, methods based on transcription activator-like effector nucleases (TALENs) or clustered regularly interspaced short palindromic repeats (CRISPRs) have been used. We have focused on the latter technique, which is based on the delivery of the Cas9 protein coupled to an appropriate single guided RNA (sgRNA) into the cells. The sgRNA guides the Cas9 protein to a specific desired DNA sequence and cuts the DNA, resulting in either Non-Homologous End Joining (NHEJ) or Homology Directed Repair (HDR). If a homologous construct or single-stranded oligodeoxynucleotide containing the corrected mutation is supplied as a template for HDR, an isogenic line can be established. Obviously, these gene-editing methods also open the door for regenerative medicine, since the repaired cells could directly be transplanted back into the patients. One example is the successful correction of the B2M and CCR5 in hematopoietic stem and progenitor cells, which showed very little off-target effects (Mandal et al., 2014). On the other hand, one needs to be aware that this system may result in off-target effects, which means that the CRISPR-Cas9 complex may cut at other places in the genome that has the same or similar homologous sequences.

Neural differentiation and disease modeling – “dementia” in a Petri dish

The subsequent step in the process towards patient-specific disease cell models is differentiation of the iPSCs (as well as their isogenic controls) into the neuronal subtypes of interest, which are the neurons affected by the disease in the patients (Figure 4). In the case of AD, the hippocampus and entorhinal cortex of the brain are the earliest affected parts of the brain, but later in the disease, wider areas of the frontal and temporo-parietal cortex are affected. Numerous protocols for differentiation of neuronal subtypes of importance for AD modeling have been published ranging from unspecific neurons to cortical neurons and basal forebrain cholinergic neurons (for review, see Freude et al., 2014). These differentiation protocols enrich for certain neural cell populations, but there tend also to be other neural subtypes including glial cells in the resulting populations. Eventually, this can be considered as an advantage since a growing body of evidence has shown that the best results in neural differentiation and maturation is achieved when different neurons and glial cells are cultured together. We have found that the maturation of the neurons into adult-like cell types, which are capable of recapitulating an AD-associated phenotype, is not trivial and poses several challenges. Hence, it may be necessary to stress the neurons by e.g. oxidative stress in order to provoke the disease phenotype.

Figure 4. Neurons differentiated from induced pluripotent stem cells (iPSCs) after immunostaining for MAP2 and Tau (two proteins stabilizing the microtubules of the neuron). Counterstained for DNA. Photo: Shuling Zhou.

The differentiated patient-specific neural cell cultures can, together with their isogenic controls, be used for modeling of AD and allow for both in depth studies of molecular disease mechanisms and, in the future, for implementation in drug discovery (Fig. 1). A number of different AD-related phenotypes have been discovered in iPSC-derived models from patients including increased Abeta 42:40 ratios, increased Tau phosphorylation and enlargement of endosomes. In line, more advanced 3D differentiation of immortalized neural progenitor cells, genetically modified to express APP and mutated PSEN1, it has recently been possible to demonstrate both intercellular Abeta plaques and intracellular tangles of hyperphosphorylated Tau (Choi et al., 2014). Finally, is has also been demonstrated that the disease phenotype in the cell models may be reverted. Studies suggest that inhibitors of beta- and gamma-secretases may revert the A-beta phenotype and GSK-3beta inhibitors can revert the Tau hyperphosphorylation. Hence, the patient-specific neural cell models may in the future be used in drug development for identification of novel targets as well as for screening of compound libraries. Interestingly, not only the familial types of AD, but also sporadic cases may be mimicked by in vitro cell cultures (Israel et al., 2012).

Cell-based therapy – the future perspective

Finally, as touched upon earlier, iPSC-derived neural progenitor cells may also pave the way for cell-based therapy. In particular Parkinson’s disease is a target for such interventions, as major symptoms of the disease is caused by the loss of a well-defined population of dopaminergic neurons located in the substantia nigra of the midbrain. More than two decades ago it was demonstrated at Lund University that transplantation of dopaminergic progenitor cells derived from aborted embryos could, at least in some cases, alleviate the symptoms (Lindvall et al., 1990). A significant problem for such therapy was the shortage of donor tissue. Interestingly, a research group at Lund University, headed by Professor HåkanWidner, will, in 2015, initiate a new round of such transplantations, and in 2018 it is planned to perform transplantation of dopaminergic progenitors derived from embryonic stem cells. However, already now, Professor Jun Takahashi at Kyoto University plan to perform transplantations based upon iPSC-derived dopaminergic progenitors as a first attempt to use the iPSC-technology for cell-based therapy of neurodegeneration.

The potentials for using iPSC-derived neural progenitor cells and neurons for disease modeling and cell-based therapy in relation to neurodegenerative disorders are great and there is no doubt that the coming years will give significant insight to this novel area of biomedicine.


The research leading to these results has received funding from the People Programme (Marie Curie Actions) of the European Union’s Seventh Framework Programme FP7/2007-2013/under REA grant agreement n° PIAPP-GA-2012-324451 (STEMMAD). Additionally, funding was obtained from the Danish National Advanced Technology Foundation grant number 047-2011-1, by the U.S. Department of Agriculture award 2011-67015-30688. Imaging data were collected at the Center for Advanced Bioimaging (CAB) Denmark, University of Copenhagen, and by the Innovation Fund Denmark (BrainStem and NeuroStem)


Choi, S.H., Kim, Y.H., Hebisch, M., Sliwinski, C., Lee, S., D’Avanzo, C., Chen, H., Hooli, B., Asselin, C., Muffat, J., Klee, J.B., Zhang, C., Wainger, B.J., Peitz, M., Kovacs, D.M., Woolf, C.J., Wagner, S.L., Tanzi, R.E. and Kim, D.Y. (2014) A three-dimensional human neural cell culture model of Alzheimer’s disease. Nature 515:274–278.

Hoeffding, M.K. and Hyttel, P. (2014): Ultrastructural visualization of the mesenchymal-to-epithelial transition during reprogramming of human fibroblasts to induced pluripotent stem cells. Stem Cell Res 14:39–53.

Israel, M.A., Yuan, S.H., Bardy, C., Reyna, S.M., Mu, Y., Herrera, C., Hefferan, M.P., van Gorp, S., Nazor, K.L., Boscolo, F.S., Carson, C.T., Laurent, L.C., Marsala, M., Gage, F.H., Remes, A.M., Koo, E,H. and Goldstein, L.S.B (2012) Probing sporadic and familial Alzheimer’s disease using induced pluripotent stem cells. Nature 482:216–220.

Lindvall, O., Brundin, P., Widner, H., Rehncrona, S., Gustavii, B., Frackowiak, R. et al. (1990) Grafts of fetal dopamine neurons survive and improve motor function in Parkinson’s disease. Science 247:574–577.

Mandal, P.K., Ferreira, L.M., Collins, R., Meissner, T.B.4, Boutwell, C.L., Friesen, M., Vrbanac, V., Garrison, B.S., Stortchevoi, A., Bryder, D., Musunuru, K., Brand, H., Tager, A.M., Allen, T.M., Talkowski, M.E., Rossi, D.J., Cowan, C.A. (2014) Efficient ablation of genes in human hematopoietic stem and effector cells using CRISPR/Cas9. Cell Stem Cell 15:643-52.

Rasmussen, M.A., Holst, B., Tu?mer, Z., Johnsen, M.G., Zhou, S., Stummann, T.C., Hyttel, P. and Clausen, C. (2014) Transient p53 suppression increases reprogramming of human fibro-blasts without affecting apoptosis and DNA damage. Stem Cell Reports 12:609-616.

Takahashi, K., and Yamanaka, S. (2006) Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126:663–676.