Rapid advancements have occurred in induced pluripotent stem cell research within the 3 years since Yamanaka and colleagues first reprogrammed adult mouse fibroblasts to an embryonic stem cell-like state by the forced expression of a small cohort of transcription factors. Progress has been made in overcoming various technical obstacles, including oncogenic threat, that hinder the application of iPS cell technology as a therapeutic strategy in humans. Remaining hurdles include the low efficiency of iPS cell induction and the demonstration of complete developmental potential. This latter impediment now appears to have been overcome simultaneously by two groups (Kristen Baldwin and colleagues and Qi Zhou and colleagues), who have generated viable adult mice from tetraploid complementation assays using iPS donor cells. The generation of sufficiently reprogrammed iPS cells and mice will allow for adequate genomic and functional testing to evaluate their utility in research applications and patient-specific cell replacement therapies, which may include infertility.
The isolation and derivation of embryonic stem cells (ESCs) from early human embryos (Thomson et al., 1998) sparked huge public and scientific interest, promising cells of unlimited potential for growth with the ability to differentiate into all cell types of the adult body. Hence, ESCs hold enormous promise for regenerative medicine and cell/tissue replacement therapies for many debilitating injuries, diseases and age-related degenerative disorders. Many stem cell transplantation treatments have been envisioned, including heart muscle cells for heart failure, cartilage for arthritis and joint injuries, pancreatic cells for diabetes, and neurons for the treatment of various neurological disorders are among the many foreseeable therapies.
The great potential of ESCs is hampered by ethical (Coglan, 2005) and immunological incompatibility issues (Baker, 2008) surrounding the generation and transplantation of human ESCs. Furthermore, the ESC field was tainted by the controversy surrounding the fraudulent claims of cloned human ESCs (Cyranoski, 2006). The promising alternative of adult stem cells was disadvantaged by perceived isolation and potency limitations. Thus the stem cell research field was poised to refocus its efforts following the groundbreaking work of Yamanaka and colleagues in indentifying key transcription factors, which reprogram differentiated somatic cells to a pluripotent state (Takahashi and Yamanaka, 2006). Ectopic expression of the transcription factors OCT4, SOX2, KLF4 and c-MYC or a different set of pluripotent factors (e.g. OCT4, SOX2, NANOG and LIN28) can induce mouse and human adult fibroblasts to take on a pluripotent ESC-like state (Takahashi and Yamanaka, 2006; Okita et al., 2007; Park et al., 2008b). These induced pluripotent stem cells (iPSCs) are similar to ESCs in morphology, pluripotency, gene expression and epigenetic profiles, and hold great promise for regenerative medicine and in vitro disease modeling, but currently possess considerable technical barriers that must be overcome before their use in therapeutic applications.
Approaches proposed to circumvent the limitations of low reprogramming efficiency include the use of less differentiated somatic cells, such as adult stem cells, instead of terminally differentiated cells and the use of small-molecule chemicals/inhibitors or growth factor pretreatments that facilitate the reprogramming process as well as novel pluripotency reporter systems to select for reprogrammed cells (Silva et al., 2008; Hotta et al., 2009; Kim et al., 2009b). Neural stem cells isolated from mice that endogenously express Sox2, c-Myc and Klf4, can be virally induced into the pluripotent state with only one (Oct4) factor (Kim et al., 2009b). Valproic acid (VPA), a small molecule histone deactylase inhibitor (Huangfu et al., 2008), and selected protein kinase inhibitors (Silva et al., 2008) promote reprogramming of differentiated cells to a pluripotent state. James Ellis' group has recently developed a novel reporter system that includes an early transposon promoter with Oct-4 (Pou5f1) and Sox2 enhancers to overcome vector silencing and to specifically mark pluripotent stem cells for easier identification and generation of iPSC lines (Hotta et al., 2009). Interestingly, expression of the reprogramming factors in somatic cells triggers senescence (Banito et al., 2009), and suppression of p53 or other senescence/apoptotic effectors greatly improves the efficiency of human and mouse iPS cell generation (Hong et al., 2009; Marion et al., 2009a). These new results suggest that the p53–p21 DNA damage response pathway acts as a barrier to iPSC generation.
Methods that curb or avoid the increased cancer risk from viral integration of transgene(s) exploit different combinations of pluripotency factors or use alternative delivery techniques that limit/prevent the integration of viral and plasmid DNA into the host chromosomes (Gonzalez et al., 2009; Zhou and Freed, 2009). One of the first steps was to replace the protooncogene c-Myc as a reprogramming factor (Nakagawa et al., 2008), which was previously attributed to causing tumour formation in chimaeric mice (Okita et al., 2007). In addition, a number of recently published studies have produced iPS cells without any permanent modifications to the host-cell genome (Okita et al., 2008; Stadtfeld et al., 2008; Kaji et al., 2009; Woltjen et al., 2009). One of these is an ingenious pluripotency gene delivery system developed by Andras Nagy's group using the piggyBac transposon/transposase method (Woltjen et al., 2009). A multi-protein expression vector incorporating all of the reprogramming genes utilizing the innovative 2A oligopeptide sequence as a spacer (Okita et al., 2008) was readily integrated into the genome. More importantly, this construct could also be efficiently removed through transient expression of the transposase enzyme, leaving the host genome unchanged. However, despite recent improvements to iPSC methodology that increase reprogramming efficiency and overcome the limitations associated with viral integration of oncogenic transgenes (Kim et al., 2009b; Woltjen et al., 2009; Yu et al., 2009), further studies are required to improve the utility of iPS cells for clinical transplantation-based therapies.
Full developmental potential of iPSCs
A remaining technical barrier in the reprogramming process is the inefficient induction of full developmental pluripotency, since all previous attempts failed to produce viable offspring from iPSCs through tetraploid complementation assays (Meissner et al., 2007; Wernig et al., 2007; Kim et al., 2008). Together with reports of reproducible gene expression differences between iPSCs and ESCs (Chin et al., 2009), and reduced teratoma forming efficiencies for human iPSCs (Park et al., 2008a; Ellis et al., 2009), these findings suggest that direct cellular reprogramming may be insufficient in restoring differentiated cells to full pluripotency. Furthermore, there are implied differences between human and mouse iPSCs, since miPSCs produce teratomas at much greater efficiencies. However, three recent studies, two of which appear in the September 3rd issue of ‘Nature’, demonstrate that miPSC lines can generate viable adult mice (Boland et al., 2009; Kang et al., 2009; Zhao et al., 2009).
The first ‘Nature’ report of viable iPSC mice was made by Qi Zhou and colleagues from the Chinese Academy of Sciences in Beijing (Zhao et al., 2009). In this study, they generated 37 miPSC lines from mouse embryonic fibroblasts, which expressed pluripotency markers with concomitant demethylation of their promoters, and demonstrated teratoma formation as well as global gene expression patterns similar to those of ESCs. Many of these lines produced chimaeric mice with germline transmission and three of these lines could produce viable, live born offspring by tetraploid complementation (0–3.5%). These iPSC lines were produced by infecting mouse embryonic fibroblasts with an Oct4-enhanced green fluorescent protein (eGFP) reporter with the four original ‘Yamanaka factors’ and then cultured under typical ESC conditions. GFP-positive clones/colonies were subsequently picked at 14, 20 and 36 days post-infection. The interesting aspect of this article was the greater developmental potential for chimeric embryos generated from iPSC lines derived from clones picked earlier in the reprogramming procedure than for clones selected later. Selected iPSC lines derived at Day 14 were competent for germline transmission in chimaeric mice at greater efficiencies compared with lines derived at Days 20 and 36. Live pups were produced only from injecting iPS cells derived from GFP-positive colonies at Day 14 into tetraploid blastocysts. Day 20 iPSCs showed superior fetal development to E13.5 than iPSCs derived at Day 36 of the reprogramming process. In future experiments, interesting correlations could be made between the expression levels of the four-pluripotency transgenes and the temporal expression patterns of endogenous pluripotency genes/reporters during iPSC generation and their subsequent differentiation in vitro, and in tetraploid chimaeras in vivo.
In keeping with the hypothesis that inappropriate expression of reprogramming genes could inhibit embryonic and post-natal development, Kristin Baldwin's group at the Scripps Research Institute designed a drug-inducible lentiviral reprogramming approach to achieve tight control of transgene expression in iPSCs and their differentiated derivatives (Boland et al., 2009). The four reprogramming factors were placed under control of an enhanced tetracycline promoter and high transgene expression was induced by the presence of the tetracycline analogue doxycycline. Mouse embryonic fibroblasts were transduced with lentiviruses containing the four-pluripotency factors and 4 days later they were exposed to doxycycline (dox) with and without VPA. ESC-like colonies appeared after five (dox plus VPA) to seven (dox only) days of dox induction (9–11 days after transduction). Colonies were picked to establish iPS cell lines on Day 14 post-transduction, with 21 colonies subcloned from the dox-plus-VPA treated cells. Prioritization of these iPS cell lines for tetraploid complementation assays involved scrutinizing them for similarities to ESCs by morphology, proliferation rate, expression of pluripotency markers and the ability to generate embryoid bodies. Included in their pluripotency-ranking repertoire was the use of a genetic marking strategy using a non-specific cell type β-galactosidase marker and a Cre-inducible cell type-specific GFP-reporter that assessed the ability of the iPSCs to differentiate into rare neuronal subtypes of olfactory bulb mitral cells in embyroid bodies and chimaeric mice. After karyotypic analysis of 12 inspected lines, three euploid lines were tested in diploid blastocyst injection assays, with all producing chimaeric mice with cells contributing to all germ layers including the olfactory bulb mitral cells. Tetraploid complementation assays carried out on selected inspected iPSC lines produced live iPSC mice at relatively high frequencies (0–13%).
A number of factors could explain the successful development of viable adult iPS cell-derived mice in these two studies. First, there were approximately 3–4-fold greater 4N injections carried out in both reports compared with the previous studies that failed to produce viable animals from iPSCs injected into tetraploid blastocysts (Meissner et al., 2007; Wernig et al., 2007; Kim et al., 2008). The longer duration of unique reprogramming conditions, and the utilization of rigorous reprogramming selection criteria would certainly contribute to full development. Particularly important is that in both studies; they ensured that all reprogramming factor transgenes were silenced (or nearly silent) in the iPSC lines tested by tetraploid complementation. Although transgene expression was not monitored during fetal development, Baldwin's research group showed that iPSC lines, which generated iPSC mice most efficiently, had reduced expression of all four reprogramming factors, whereas the less efficient line had detectable expression of Klf4 and/or Oct4 in the absence of dox.
The generation of viable mice through tetraploid complementation provides definitive evidence that some iPSCs are in fact truly pluripotent. However, as with the somatic cell nuclear transfer field (Niemann et al., 2008), the reasons why only a small percentage of differentiated somatic cells can be reprogrammed to full pluripotency are not fully understood. The answer probably lies, in part, with inducing the appropriate temporal and spatial reactivation of the endogenous pluripotency transcriptional network. Another possibility could be the loss, or alterations in, various epigenetic modifications, which are necessary to maintain genomic imprinting in the donor somatic cells. Further knowledge in these areas will undoubtedly improve cellular reprogramming efficiencies. But do we really need to have full cellular reprogramming? Some iPSC lines have a tendency to differentiate into particular lineages over others depending on their somatic cell source (Cheryle Séguin, personal communication). These lines will undoubtedly help elucidate the factors necessary to guide specific cell fate programmes. Furthermore, the lines could also be used to enrich therapeutic cell populations for the treatment of certain diseases or injuries.
Proof-of-principle studies have demonstrated that iPSCs can be used as transplantation treatments in various mouse models of human diseases in vivo (Hanna et al., 2007; Wernig et al., 2008). In addition, numerous pluripotent stem cell lines have been generated from the cells of human patients with specific diseases (Dimos et al., 2008; Park et al., 2008a; Ebert et al., 2009; Lee et al., 2009; Maehr et al., 2009; Soldner et al., 2009; Ye et al., 2009). These in vitro disease models will undoubtedly create a wealth of new knowledge for various ailments but may also be used as drug screening devices for the development of new therapies. Ultimately, translating iPSC technology from the lab bench to human clinical transplantation applications will require direct reprogramming of differentiated cells and/or adult stem cells without transgenesis. A combined approach is most likely, using an optimal donor cell source, innovative pluripotency factor delivery methodologies and optimized cell culture conditions. Here, the study of stem cells from reproductive tissues has already made a potentially significant contribution. Amniotic-derived and umbilical cord blood-derived stem cells endogenously express some of the pluripotency markers and their differentiation potential can be broadened by long-term growth factor exposure (Tamagawa et al., 2004; Rogers et al., 2007; You et al., 2008; Trovato et al., 2009; Betts, unpublished). These somatic stem cells easily collected before, or at the time of birth, could become a highly favourable cell source for patient-specific iPSC generation (Giorgetti et al., 2009; Haase et al., 2009; Li et al., 2009). Pluripotency induction from cord blood or amniotic cells would allow ample time for cell expansion of the iPSCs themselves but also their differentiated cell/tissue products that could be propagated, cryopreserved and banked for rapid ‘off-the-shelf’ therapies in the future. Recently, human iPSCs have been generated, albeit at a much lower efficiency, by the direct delivery of recombinant reprogramming proteins anchored to cell penetrating peptide sequences (Kim et al., 2009a; Zhou et al., 2009). Moreover, hypoxic or low oxygen (5% O2) culture conditions induce and manipulate endogenous expression of stem cell genes in somatic cells without genetic manipulation (Page et al., 2009), and hypoxia significantly increased the efficiency of iPSC generation by non-viral means (Yoshida et al., 2009). Low oxygen tension also induces telomerase activity (Betts et al., 2008), a factor known to change the cell state towards a progenitor stem cell-like condition (Perrault et al., 2005) and has been used as a pluripotency factor in the production of iPSCs (Marion et al., 2009b) and in the maintenance of ESCs in the undifferentiated state (Yang et al., 2008). Boosting the advantages of hypoxia, is that physiological oxygen has the secondary benefit of lessening free radical formation and reducing the number of subsequent mutations (Forsyth et al., 2006). Studies like those above are establishing a role for the hESC/hiPSC-derived microenvironment in the regulation of self-renewal and fate (Bendall et al., 2007; Peerani et al., 2007).
Nevertheless, the results presented in the two papers by Boland et al. and Zhao et al. suggest that iPSCs are equivalent to ESCs in pluripotency and developmental potential and will provide a new resource to examine the functional and genomic stability of cells and tissues derived from iPSCs. However, caution is still urged on the use of ESCs/iPSCs in therapies even after the recent approval this year by the US Food and Drug Administration (FDA) for the first clinical trial to use derivatives of human ESCs (Barde, 2009). Clearly, iPSCs will circumvent any possible immune rejection of transplanted precursor cells derived from ESCs. However, there is no reliable information about the potential of ESCs/iPSCs, or their differentiated derivatives, to de-differentiate in situ after transplantation. Furthermore, their ability to undergo tissue-specific differentiation and produce functional cell types using protocols optimized for hESC differentiation is untested. The importance of the microenvironment in the regulation of cell fate is demonstrated by the ability of ESCs/iPSCs to differentiate in a niche-dependent manner. When injected into immune-compromised mice, hESCs form differentiated teratomas (Thomson et al., 1998; Ludwig et al., 2006), but when injected into human tissue engrafted into mice, hESCs form primitive undifferentiated tumours, a difference attributable to the presence of human factors that are not present in mice (Shih et al., 2007). Fundamental knowledge of the necessary factors required to differentiate ESCs and iPSCs towards each specific cell lineage must be obtained.
Important issues need to be resolved before human ESC/iPSC-derived cells or tissues can be employed as a cell/tissue replacement strategy and contribute to the treatment of degenerative diseases in a clinical setting. These issues include the assessment of therapeutic outcome and safety in clinically relevant animal models, especially in non-life threatening situations that are quality of life issues (e.g. joint diseases, infertility) which will necessitate many layers of long-term safety and efficacy evaluations. Although useful in providing proof-of-principle results, significant anatomical, physiological, disease/injury presentation and clinical response differences between humans and rodents seriously limit the use of the mouse for these evaluations (Patterson, 2000; Starkey et al., 2005; Koch and Betts, 2007; Koch et al., 2009). Although a handful of promising therapeutic outcomes have been obtained in rodent models (Lanza et al., 2004; Chang et al., 2006; Capoccia et al., 2009), hESCs differ significantly from mouse ESCs (Ginis et al., 2004; Pera and Trounson, 2004). Large animal models that represent important aspects of human anatomy, physiology and pathology more closely than mouse models are urgently needed for studies evaluating the safety and efficacy of stem cell therapies (Koch and Betts, 2007; Koch et al., 2007; Wilcox et al., 2009). A move from concept to cure in a naturally occurring animal disease or injury (not artificially induced in small laboratory animals) will not only help develop strategies for human regenerative medicine but will also provide the leverage needed to increase public support and acceptance of stem cell research and therapy in humans (Fiester et al., 2004).
iPS cells represent a potential revolution in our understanding of human reproduction. At the level of fundamental knowledge, iPS cells offer a valuable in vitro model to study the formation of various human reproductive cell types. They will be of potential value in studying human gametogenesis and epigenetic modifications of the germ line. In the mouse, immature sperm cells derived from mouse ES cells in culture have produced live offspring (Nayernia et al., 2006). The possibility that human iPS cells could be induced to form germ cells is tantalizing. Future treatments for infertility could employ iPSC-derived germ cells that are genetically identical to the patient having been obtained from the patient's own tissues. Alternatively, these patient-specific iPSC lines could be utilized as cell models to examine cases of unexplained infertility. Although the use of iPSC technology as a infertility treatment would be considered by some as less ethically challenging than using somatic cell nuclear transfer (reproductive cloning) (French et al., 2008), the expected increase in epigenetic alterations during prolonged in vitro gametogenesis (Eppig and O'Brien, 1996; Eppig and O'Brien, 1998) would warrant extensive testing before it's clinical application as a fertility tool.
Studies from D.H.B.'s laboratory discussed herein and in the preparation of this article were supported by the Canadian Institutes of Heath Research (CIHR MOP-86453) and the Natural Sciences and Engineering Research Council of Canada (NSERC 250191-07). B.K. is supported by the National Health and Medical Research Council of Australia (Grant No. 509178).
Drs Cheryle Séguin and W. Allan King are thanked for their intellectual insights and critical reading of the manuscript.