S. D. Qi MBBS, BMedSci; P. D. Smith MBBS, PhD, FRACS; P. F. Choong MD, FRACS, FAOrthA.
Induced pluripotent stem cells (iPSCs) are generated from somatic cells by the exogenous expression of defined transcription factors. iPSCs share the defining features of embryonic stem cells (ESCs) in that they are able to self-renew indefinitely and maintain the potential to develop into all cell types of the body. These cells have key advantages over ESCs in that they are autologous to the donor cells and can be generated from individuals at any age. iPSCs also circumvent ethical and political issues surrounding the destruction of embryos that is necessary in the isolation of ESCs. This review briefly describes the advent of iPSC technology and the concepts of nuclear reprogramming, and discusses the potential application of this powerful biological tool in both surgical research and regenerative medicine.
Stem cells have generated considerable interest in the public and research. The cloning of Dolly the sheep and isolation of human embryonic stem cells (ESCs), in particular, have led to the investigation of stem cell-based regenerative therapies and provided insight into the developmental basis of disease. However, such technologies have often been hindered by the ethical and political dilemma that surround the use of oocytes and the destruction of human embryos needed to isolate these cell types. Induced pluripotent stem cells (iPSCs) are somatic cells that have been reprogrammed to an ESC-like state via the exogenous expression of transcription factors. Not only do they bypass the need for embryos or other ethically charged cells, they are autologous and can be generated from individuals of any age. iPSCs have provided a novel and unique tool for both understanding disease and regenerative medicine. For surgeons, iPSC offers an unlimited source of cells that may not be normally available in the patient and amenable to correction of genetic defects ex vivo.
Properties of a stem cell
A stem cell is broadly defined as a cell that is capable of self-renewal and generating several differentiated cell types and can be considered as a functional unit of embryogenesis and adult tissue regeneration. Stem cells can be broadly categorized into two types: embryonic and adult. Much attention has been given to ESCs, a type of pluripotent stem cell derived from the inner cell mass of blastocysts, due to their potential application to regenerative medicine. ESCs are distinguished in their ability to give rise to cells of all three germ layers of the body and self-renew indefinitely under certain defined conditions.
The concept of a stem cell started with Virchow proposing that all cells are derived from other cells in 1855. A ‘stem cell hypothesis’ was later proposed in 1908 by Alexander Maximov as a means to explain the regenerative capacities of haematopoietic cells. It was not until 1963 when the studies of Till and McCulloch on the haematopoietic stem cell that the concept was truly established. Currently, Melton and Cowen proposed that a complete description of any stem cell involves evaluation of three key cellular properties: potency – the developmental potential of a cell; exceptional replicative capacity; and clonality – the ability of a cell to generate an entire organ or individual.
Development and the epigenetic landscape
Mammalian development involves a unidirectional progression of cells from higher potential towards terminally differentiated cells to fulfil specialized phenotypes of somatic tissues.[7, 8] With the exceptions of lymphocytes that undergo gene rearrangement, all somatic cells within an organism share an identical genome (sequence of DNA) – cell types are determined by their unique gene expression profiles. Epigenetic mechanisms, which do not change the sequence of DNA, regulate gene transcription by inducing stable changes in gene expression by altering the accessibility of the chromatin state. The establishment of such gene expression patterns is set and maintained by transcription factors via histone modifications and DNA methylation and is largely an irreversible process, with the exception of germ cells.
A proposed mechanism for understanding the various states of potency is the epigenetic landscape by Waddington (Fig. 1). A cell's genetic expression profile, and subsequently its identity, is depicted by a marble rolling down the surfaces of hills and valleys that represent various stable and unstable cellular phenotypes, analogous to the potential energy states of matter described by chemists. The vertical height of each point on the surface represents the cell fate potential, whereby the gradient and surrounding contours at each point can be thought of as a force either pushing the cell towards a new fate (rolling down a hill) or maintaining a stable fate and providing barriers that need to be overcome in order for the cell to progress to another fate (in a well). What is also reflected in the model is that, as the marble rolls down the hill, a cell loses its potency, and for it to return to a higher potency state, it should present a much higher barrier than if it were to proceed further down the landscape. Therefore, a pluripotent stem cell would be depicted high up in the landscape, capable of rolling down towards the bifurcations into any of the valleys representing terminally differentiated cells, gradually losing their ability to choose other paths (cell fates).
The term nuclear reprogramming is used to describe either functional or molecular changes to cells undergoing fate changes. When used as a functional term, reprogramming refers to experimentally induced, stable changes in cell fate. Nuclear reprogramming is able to revert differentiated somatic cells back into a pluripotent state with greater potential in cell fates, as would be observed earlier developmentally. There are four major techniques through which this is achieved – nuclear transfer (NT), cell fusion, cell explantation and reprogramming by defined factors (Fig. 2). Somatic cell nuclear transfer (SCNT), where a somatic cell nucleus is introduced into an enucleated oocyte, has been the classical method and is the only method so far by which totipotent cells (capable of producing both the embryo proper and the extraembryonic tissues) can be produced, and is the technique through which Dolly the sheep was cloned. Although considered as the gold standard, NT had significant drawbacks, preventing its translation into clinical therapies. The requirement for human oocytes, a rare and ethically debatable cell type, is further compounded by the extreme technical challenges of the method that is heavily burdened with low efficiencies. Cell fusion and cell explantation also faced equally worrisome problems of tetraploidy and cells that carried male germ cell imprinting, respectively.
Reprogramming by defined factors to generate iPSC is a reliable and convenient method by which autologous pluripotent cells can be created, revolutionizing stem cell research. The generation of iPSCs results from the culmination of three key scientific principles: the demonstration that differentiated cells retain the same genomic information as early embryonic cells; the development of culture techniques for pluripotent cell lines; and the observation that transcription factors are key determinants of cell fate. iPSCs were first generated by Shinya Yamanaka in 2006, where 24 candidate genes that had been previously shown to be involved in maintenance of the pluripotent state were screened and systematically reduced to four – octamer-binding transcription factor 4 (Oct4), sex-determining region Y box-2 (Sox2), Krüppel-like transcription factor (Klf4) and myelocytomatosis viral oncogene (c-Myc) were able to faithfully reprogramme mouse fibroblasts back into a pluripotent state. Although initially the first generation iPSCs showed many differences from the ESC counterparts, the technique has been improved upon by several other groups. To date, the latest generation of iPSCs resemble ESCs by all phenotypic and functional assessment of pluripotency (Table 1), including the most stringent of tetraploid complementation to demonstrate that iPSCs are able to generate a live individual. Furthermore, it has been demonstrated that aberrant silencing of imprinted genes on chromosome 12qF1 can distinguish between iPSCs that have been faithfully reprogrammed when compared with the gold standard of SCNT. The four reprogramming factors have shown to be universal across various cell types and species, suggesting that the reprogramming process is likely a universal one.
Table 1. Comparison of embryonic stem cell (ESC) and induced pluripotent stem cell (iPSC). Tests of pluripotency and their relative levels stringency, with morphology being the least stringent and functional assays being the most stringent[13, 15-17]
Propensity to differentiate towards particular cell types
ESCs have been documented to show variability in differentiation yields between different lines
iPSCs may demonstrate ‘epigenetic memory’ – a propensity to differentiate more easily into the cell type they were derived from, but this seems amenable to passage and manipulation
Many tissues in the human adult are unable to regenerate after disease or injury. Notably, tissues such as that of the central nervous system are commonly affected by diseases for which there is no current cure, as the tissues have minimal capacity to regenerate once they have been destroyed. Transplantation of tissues to ameliorate surgical removal of diseased organs faces several challenges. Donor tissues are often limited by supply, by the number of allogeneic donors (e.g. kidney) or limited by potential harm to another site within the patient (e.g. skin grafts). The damaged tissue may not have a suitable counterpart in the adult patient, such as progenitor populations.
ESCs have been considered to be a potential source for cellular replacement therapies due to two properties: the ability to renew indefinitely in culture and thereby provide a large source of cells and their versatility in being able to differentiate into any cell type needed. However, as the isolation of ESCs requires the destruction of discarded embryos, there exists great ethical and political debate on their use. Furthermore, all cell therapies that involve ESCs would ultimately face the same technical challenges of conventional allografts,[20, 21] that is, immune rejection.
Autologous stem cells have key advantages to cellular replacement therapies as they would circumvent ethical and political burdens. They would also avoid the use of immunosuppressant therapies, which have significant side effects such as hypertension, increased risk of infection, as well as having been shown to directly interfere with both normal and transplanted tissues such as inhibiting the replication of the beta cells of the pancreas.
Inherited genetic defects that carry through to iPSCs present a challenge but can be tackled one of two ways: either through the use of human leukocyte antigen (HLA) matched allogeneic cells or via gene therapy[25, 26] (see Fig. 3). Gene therapy methods, such as lentiviral vectors or homologous recombination, applied in vitro also means that the patient does not have to be exposed to any risks associated in vivo, such as exposure to the viral vectors. The resulting cells may be controlled for quality prior to introduction to the patient, for example, by selecting a purified cell population through flow cytometry and undergoing rigorous screens for genetic abnormalities.
Various protocols have been established for the differentiation of either ESCs or iPSCs into a certain desired cell type. Of note, protocols exist to generate motor neurons,[28, 29] photoreceptors, osteoblasts, cardiomyocytes, beta islet cells, hepatocytes and intestinal tissue. See Table 2 for cell types derived from iPSC.
Table 2. Examples of cells derived from pluripotent cells (embryonic stem cell/induced pluripotent stem cell (ESC/iPSC)) with defined protocols and iPSC-derived disease-specific cells. Adapted from[17, 28, 30-32, 34-44]
Cell types derived from iPSC
CNS, central nervous system; VPA, valproic acid.
Long QT type 1
Long QT type 2
E-4031 and cisapride aggravate disease phenotype, nifedipine, pinacidil and ranolazine ameliorate some aspects of disease phenotype
Lesch-Nyhan syndrome (carrier state)
Type 1 diabetes
Glycogen storage disease 1a
Skeletal muscle cells
Hereditary tyrosinaemia, type 1
Gaucher's disease, type 3
Alpha-1 antitrypsin deficiency
Amyotrophic lateral sclerosis
Motor neurons, glial cells
Spinal muscular atrophy
Neurons, astrocytes, mature motor neurons
VPA and tobramycin ameliorate phenotype
Transplanted neurons ameliorate phenotype in rat model
IGF1 and high dose gentamicin treatment ameliorates phenotype
Loxapine improves neuronal connectivity, no improvement with clozapine, olanzapine or risperidone
Sickle cell anaemia
Duchenne muscular dystrophy
Becker muscular dystrophy
Retinal progenitors, photoreceptor precursors, retinal-pigment epithelium cells, rod photoreceptors
α-Tocopherol ameliorates disease phenotype with certain backgrounds.
Many proposed models have already been supported with proof-of-principle experiments in animal models. A model of sickle cell anaemia in mouse has been shown to be cured by a combination of iPSC and homologous recombination, whereby the gene for sickle cell is corrected in iPSC derived from the skin and then redifferentiated into a haematopoietic stem cell and reintroduced into the mouse. This has also been achieved with Fanconi's anaemia. Alpha-1 antitrypsin deficiency has also been amenable to the combination of gene therapy and iPSC technology to generate autologous disease-free hepatocyte-like cells. iPSC-derived photoreceptor precursors have been derived and introduced into mouse models of retinal degeneration to demonstrate integration and efficacy at restoring function.
How generated cells will integrate back into the host is not an issue to be overlooked. It may explain why haematopoietic diseases have been the first to be targeted via the iPSC approach. There may exist changes at the site of pathology that prevent the integration of new tissues. The orientation and three-dimensional structure of the regenerated tissue are also important. An advantage of deriving a cell for transplantation from an embryonic state would be that it is possible to control what level of differentiation the cell is at, so as to induce a cell state that would be optimum for integration, rather than fully mature phenotypes. Cells may also be integrated into biodegradable polymer scaffolds, which can provide structure or elute drug molecules to enhance integration or the natural repair process.[31, 51, 52] The introduction of various cell types will require tissue-specific strategies. An important factor is that the cell products do need to be differentiated satisfactorily to avoid the introduction of undesired cell types, or developmentally premature cell types that may give rise to immunogenic responses to products not normally present in adult patients, or more worryingly, malignant transformation.
Disease modelling and drug development
The current study of many human diseases is achieved by using animal models and cell lines that approximate the behaviour of certain cell types, as the accessibility of many disease-afflicted tissues is limited, such as with the retina. As iPSCs grow indefinitely in culture, they provide an unlimited source of the desired cells with little harm to the patient. They are also capable of differentiating into other cell types that may be involved in the disease process.
The advantage of autologous pluripotent cells is that they have the ability to create more accurate disease models, as well as provide insight into the mechanistic basis of the disease. Although there is success in modelling of diseases in vitro by introducing genetic defects known to be responsible into ESC,[35, 53, 54] such models are only limited to mechanisms of disease that have direct genetic causations and do not capture the disease on the patient's own genomic background. For example, fewer than 10% of amyotrophic lateral sclerosis (ALS) patients are known to suffer from the familial form. Autologous models reflect the individual response of the patients at concern in the context of their own particular genetic identity, including sporadic forms of disease and those affected by complex multifactorial diseases of unknown genetic identity, such as autism spectrum disorders and type 1 diabetes.
Disease modelling using pluripotent cells would also allow the researcher to observe earlier stages of the development of disease, as disease has progressed considerably past the point of first onset at the time of diagnosis, or even further back in development. There have been a number of successfully generated patient-specific iPSC lines and subsequent disease modelling, such as motor neurons from an 82-year-old ALS patient, motor neurons from spinal muscular atrophy (SMA) and familial dysautonomia – in which it was demonstrated that there were specific defects in neurogenesis and the migration of neural crest precursors, tissues that were previously unobtainable.
Being able to recapitulate diseases accurately in vitro will facilitate the discovery of new therapeutic agents, which can either be deployed for the disease in general or be used to characterize individual pharmacokinetic responses. Pharmacogenetics is a growing field, but as only a few genes predictive of responses have been identified, iPSC-derived models of disease provide both an effective and a functional assessment of how an individual will respond to a certain therapeutic agent. The use of this tool in high-throughput screening assays could allow better prediction of the toxicology caused by new drugs and offer insight into underlying mechanisms. In particular, hepatotoxicity and cardiotoxity are two principal causes of drug failure during pre-clinical testing, which can be tested in vitro for individuals through the derivation of these cells even when they are not involved in the particular disease process. Further benefits of this approach would be the reduction of substantial risk and cost involved in early stage clinical trials. Three reports of iPSC derived from patients suffering from SMA, familial dysautonomia and LEOPARD syndrome not only recapitulated the cell abnormalities in vitro, the symptomatic phenotype was partially alleviated when exposed to experimental drugs in culture.
iPSC and cancer
In addition to understanding particular diseases, attempts to understand the mechanisms behind reprogramming have also shed light into biological development and cancer. Signalling pathways that have been shown to significantly influence reprogramming efficiency are also known to have active roles in cancer. Of note, it is recognized that several tumour suppressor proteins such as p15, p16, p19 and p53 are roadblocks in the reprogramming process.[58, 59] Knockout of p53 in mouse cells leads to significant increases in reprogramming efficiency and faster kinetics.
Two recent reports have shown that a mesenchymal-to-epithelial transition is a critical event during the derivation of iPSCs from fibroblasts, a process that is pivotal during organ development and its reverse epithelial-to-mesenchymal transition in cancer metastasis by endowing cells with migratory and invasive properties, respectively.
Transdifferentiation involves the transformation from one cell type to another directly, without reversion to a pluripotent state. Many experiments prior to the advent of iPSC have demonstrated that it is possible to induce fate switching with the introduction of the correct combination of key factors and these experiments, such as MyoD1 paved the path towards the creation of iPSC. The success of iPSC has renewed interest in finding the transcription factors capable of converting mature cell types directly into each other. Progress has already been seen in both in vitro and in vivo models.
For clinical applications, transdifferentiation may be more advantageous than reverting a cell back to pluripotency and then back to a differentiation pathway, as with the right initial cell type, it could both increase efficiency as well as decrease the risk of tumourigenicity of the resulting cells. It has been shown that this not only possible between highly related cells such as B cells and macrophages, but is also possible both across germ layers, as mesenchymal fibroblasts have been transdifferentiated to functional ectodermal neurons, including subtypes of neurons that are able to achieve integration, as well as endodermal hepatocytes capable of rescuing liver function in live mice. What has also been demonstrated elegantly is that reprogramming can be achieved in vivo, as shown by Melton's group where mouse pancreatic acinar cells have been transformed to functional beta islet cells capable of relieving a hyperglycaemic state, through the viral introduction of the transcription factors Ngn3, Pdx1 and Mafa.
One of the most exciting features of iPSC is the fact that they circumvent the ethical and political dilemma of ESC by bypassing the need to destroy embryos. However, this is not to say that iPSCs are ethically burden-free, nor to say that ESC research has become obsolete.
The convenience by which iPSCs are created is one source of concern as this poses risks to copyright issues over an individual's own genetic sequence. Who do iPSCs belong to once they are generated – the donor of the cells or the individual who created the line? Further questions will be raised in the iPSC field when technology advances in generating certain cell types. Should infertile individuals be permitted to generate zygotes genetically identical to themselves in the name of fertility treatment? As technology in this field advances, there will be debate regarding to what extent we can ‘clone’ for therapeutic purposes.
Challenges and future directions
There still remain a number of roadblocks before we see the direct translation of iPSC technologies into the clinic.
Safety of iPSCs and derivative cells are still not fully determined. Ideally, cellular replacement therapies would like to see the complete elimination of exogenous material. The risk of tumourigenicity is a major concern arising from the use of oncogenes and their products along with the potential presence of undifferentiated cells. Two of the reprogramming factors, c-Myc[66-68] and Klf4,[69, 70] have been known to have oncogenic properties. Many of the mice transplanted with iPSC-derived cell types, or that have been generated via tetraploid complementation, have a high susceptibility to developing tumours. It has been shown that generation of identical iPSC can be achieved in the absence of these factors, either without c-Myc or without both c-Myc and Klf4. The omission of c-Myc has significantly decreased susceptibility to tumours in its chimeras.
Already there has been considerable progress towards eliminating exogenous material (Table 3), through both non-integrative (plasmid and adenovirus vectors) and excisable vectors (piggyBac transposon and CreLoxP system) or non-DNA-based vectors (proteins). Most recently, it has been demonstrated that it is possible to reprogramme cells using mRNA as well as miRNA at higher efficiencies.[74, 78] Since these developments, the majority of recently published studies have been shifting towards using non-integrative methods of generating iPSC. Integration-free iPSC kits are readily commercially available.
Table 3. Summary of current methods of generating induced pluripotent stem cell (iPSC). Adapted from[7, 17, 74-77]
Reasonable efficiency, transduces replicating and non-replicating cells
Reasonable efficiency; allows for control of factor expression
Reasonable efficiency; integrated material is removed
Reasonable efficiency; integrated material is removed
Multiple integration sites, viruses may be incompletely silenced, viral infection, slow reprogramming kinetics
Multiple integration sites, viruses may be incompletely silenced, viral infection
Multiple rounds of transfection needed; potential for immune reaction; expensive to produce
Lower efficiency compared with other methods
Extremely low efficiency, requires large quantities of protein and multiple rounds of application
Not able to completely replace all factors – still relies on transduction of at least one factor by other means, small molecules may potentially interact with other targets
Sequence-sensitive RNA replicase, difficult to eradicate virus from infected cells
Another issue that raises concern is the immunogenicity of the cells generated. As the cells are autologous, theoretically there should be no immune reaction from the host towards any iPSC-derived tissues. However, one cannot be certain that fetal or novel proteins generated through either the reprogramming or differentiation process are completely immunotolerant. For example, it has been shown that Oct4-specific T cells can be detected in >80% of healthy individuals, which would mean that iPSCs and their derived cells may generate an immune response if the Oct4 protein involved in the generation process is not completely eliminated. The use of xeno products in culturing iPSC-derived cells also need to be avoided if cells are being derived for introduction into patients.
The cost involved in generating and maintaining an iPSC line from an individual is expensive and the process is time-consuming. The current efficiency by which iPSCs are generated is low, with reports ranging from 0.001% to 4%.[7, 74] In a clinical setting, it would be largely impractical to generate individual cell lines at their current costs. There has been an initiative to bank iPSC lines based on HLA typing, which would reduce the need to generate a cell line for every individual as well as provide allogeneic donors in cases of inherited diseases.[25, 80]
The final concern of iPSC is the need for establishment of quality control of all iPSC cell lines. It is difficult to determine fully whether iPSCs generated from one laboratory are identical to those from another and whether they have all been truly reprogrammed. There is increasing evidence to show that iPSCs retain some form of ‘epigenetic memory’ from the cell type they were derived from,[15, 16] which could benefit downstream differentiation into certain cell types through strategic choice of initial somatic cell type for induction. Most recently, iPSCs have also been demonstrated to display subtle differences from their ES counterparts on chromosomal, genetic and epigenetic levels, such as aneuploidy, single-point mutations and altered methylation pattern variations between lines.[81-83] Incomplete methylation has been linked to the causation of residual epigenetic memory of reprogrammed cells, while other reports describe differentiation protocols that are unaffected by inter-line variability. Epigenetic memory may, in fact, be beneficial when wishing to increase the yield of a particular cell type during differentiation. The exact long-term functional consequences of these subtle differences are not yet known, but is the subject of an ongoing investigation.[85-87] Ultimately, it is hoped that universalized standards, based on their intended uses, would be applied to all cell lines generated. Most recently, there has been establishment of both genome-wide reference maps of DNA methylation and gene expression, along with a set of functional tests to characterize differentiation ability.[27, 88]
One of the ongoing questions has been whether they truly are equivalent to their ES counterparts. Although iPSCs resemble their ES counterparts through the most stringent tests of pluripotency, it is important to remember that these cells are likely to harbour subtle differences because of their distinct origins and modes of derivation. Considering their use both in research and their clinical applications, it should be a priority to understand their inherent properties rather than ask whether iPSCs are identical to ES in every facet.
Since the first description of iPSC generation six years ago, significant progress has been made in both improving the technology itself and defining useful clinical applications. It offers a powerful tool to individualize medicine in terms of understanding disease, predicting pharmacological responses and may lead to the hope of regenerative medicine without the need for immunosuppression. Although there are still questions to be answered and challenges to be overcome, iPSCs have created new directions in stem cell research, altering fundamental ideas about cellular identity and stimulating new progress towards personalizing medicine.
SDQ would like to thank Professor Konrad Hochedlinger for his wonderful series of lectures and support during her pre-doctorate fellowship.