In 1997, the birth of Dolly indicated that activating transcription factors in egg cells can induce the state of totipotency in somatic cells (1). This groundbreaking discovery changed our previous thought about reprograming of somatic cells, also, marked a mile stone in the field of cloning of animals. Up to now, there are reports of human cloned blastocysts following SCNT with adult fibroblasts (2).
Afterwards, James Thomson (3) established the first human embryonic stem (ES) cell line, which opened a new chapter of intense research on stem cells. Human embryonic stem cells can be used to cure such diseases as Parkinson's disease and juvenile-onset diabetes mellitus. However, for ethical and practical reasons in many primate species, including humans, the ability of ES cells to contribute to the germ line in chimeras is not a testable property. Moreover, the use of totipotent embryo, which is considered as a human being, received most severe criticism. Thus, the research on human embryonic stem cells provoked great criticism and has been banned in many countries.
In 2006, Yamanaka's lab reported that they have induced pluripotency in mouse fibroblasts by defined factors including Sox2, Oct4, Klf4, c-Myc (4). Their work demonstrated that under certain circumstance, the control of genetic expression is essential in changing their state of differentiation. More Recently, it was reported that the generation of one-factor human iPS cells from human fetal neural stem cells by ectopic expression of OCT4 alone (5). It is assumed that iPS cells have great advantages over ES cells. First, it can circumvent ethical problems and immune rejection. Second, it has a broader application: for studying cells' differentiation and comparing difference between normal cells and diseased ones; rendering drug screening in human cells possible; making regenerative medicine, which can be applied in clinical use. For all the aforementioned reasons, iPS cell enjoys tremendous success and attracts intense interest.
However, is iPS cell the panacea? There are some underlying problems worth our consideration. In fact, we know little about the origin and process of reprograming. Also, the iPS cell seems to be against the theory of natural selection. All previous studies have shown that gene expression profile of human iPS cells is comparable, but significantly different from, with that of human ES cells. The studies of the reprograming recalcitrant genes clearly indicate that iPS cell reprograming is not perfect (6). Functional difference raised from different gene expression profile of human iPS cell technology has not been investigated. However, this difference should not be underscored. Compared with physiological ontogenesis, the method of human iPS cell technology lacks the process of natural selection. Moreover, the safety of iPS cell is casted into doubt for the possibility of tumorgenesis. Finally, the claim about advantages of iPS cell in circumventing immune rejection and ethical problems is flawed.
THE ELUSIVE ORIGIN OF IPS CELL
When the first iPS cell line was established, the efficiency of reprograming was quite low. Shortly after the establishment of iPS cells in Yamanaka's lab, many independent labs successfully established iPS cells through different protocols (4,7–11). However, the low efficiency functions as a main obstacle to inducing pluripotency, which ranges between 0.01% and 0.2%. Yamanaka and coworkers (12) give three alternative explanations to this problem: “First, the origin of iPS cells may be undifferentiated stem or progenitor cells coexisting in fibroblast culture. Another possibility is that retroviral integration into some specific loci may be required for iPS cell induction. Finally, minor genetic alterations, which could not be detected by karyotype analyses, or epigenetic alterations are required for iPS cell induction.” This low efficiency is approximate to the percentage of progenitor stem cells in tissues, which is about 0.067% (13). Therefore, arguments arise that the preexisting progenitor cells can be a plausible explanation to the origin of iPS cell (14). Additionally, it is also stated that the rare adult progenitor cells presenting in the somatic tissue, which are similar to the iPS cells in DNA methylation state, may perform as the source of iPS cells (15).
Surprisingly, before substantial evidence provided by their group, Yamanaka dismissed the possible origin of the iPS cell being the rare progenitor stem cells, which exist in human tissues (16). Yamanaka's claim primarily rests on the fact that the efficiency of reprograming can reach up to 10%, which is much higher than the percentage of tissue stem or progenitor cells. However, this claim is flawed in two aspects. First, just as Yamanaka confessed: “reprogramming factors or chemical treatments might preferentially enhance the proliferation of tissue stem or progenitor cells” (16). More importantly, up to now, there is no formal criteria for identifying iPS cells. For the evaluation of the efficacy of the human iPS cell generation, the expression of Nanog, Oct4 or alkaline phosphatase (ALP) genes were used as the criteria of the iPS cell colonies (17). However, uniformity of Nanog or ALP positive colonies has not been shown. By comprehensive analyses of primary human iPS cell colonies, it has been identified that a large number of the primary human iPS cell colonies that expressing Nanog and alkaline phosphatase genes are lacking the expression of several pluripotent genes (15). The presence of partially reprogramed developmental genes in the Nanog and/or ALP positive primary human iPS cell colonies indicate that increase of the efficiency of the pluripotent cell generation cannot be concluded. Moreover, some of such markers may be not as reliable as they seem to be. For example, reports demonstrated that Oct4 expression is not required for somatic stem cell self-renewal (18), thus echoed the concern that Oct4 may not be a reliable stem cell marker (19). If on the basis of false stemness markers, the relatively higher efficiency cannot justifiably dismiss the possibility of progenitor cells or tissue stem cells. In fact, a recent research has provided the evidence that stem cells and progenitor cells are the suitable origin of mouse iPS cells (20).
In sum, before a full understanding of the process of reprograming, the origin of iPS cell will remain to be determined, and further study that shed light on it will definitely contribute to making reliable iPS cell.
AGAINST NATURAL SELECTION THEORY
Optimization of the differential stage of somatic cell source of iPS cells is not sufficient to generate healthy pluripotent stem cells, as the iPS cell method violate the theory of natural selection at the molecular level. Although we produce iPS cells through molecular techniques, we cannot guarantee that these iPS cells can carry the right genomic information. The capacity of iPS cell lines to form tumors is an unexpected byproduct of the reprograming process, perhaps arising from the erroneous genomic information transmitted without the natural selection (21,22).
In the physiological condition, only chromosomes in germline cells inherit to next generation. In this context, somatic cells are disposable. Chromosomal DNA in germline cells are subjected to a postfertilization reprograming and a primordial germ cell reprograming during ontogenesis. It is widely acknowledged that during the normal reproduction process, organisms will have mechanisms of investigation and supervision on germ cells at different stages, to guarantee the quality of the offspring. Such mechanisms include the screening in the development of eggs. Take human beings as an example, an adult woman can discharge 400 eggs or so from the ovary during her lifetime. In contrast, the number of oocytes in her ovary may reach up to 5–7 million. Moreover, the mechanism will play more important role in screening sperms in male human beings because the number of sperm is much more tremendous than oocytes. Every offspring can be treated as a miracle because only through the intense screening during the process of fertilization, can one sperm fertilize an egg cell, and thus produce an offspring. In spite of such mechanisms, the possibility of malformation maintains a high percentage. Artificial reprograming including iPS cell technology only mimics a postfertilization reprograming. It has been shown that epigenetic state of somatic cells and germline cells are significantly different (23). Accordingly, the variety of abnormal phenotypes including placenta overgrowth, large fetus syndrome, immune dysfunctions, and a shorter life span were observed in cloned animals (24). The gene expression profiles of each organ in cloned mice are extremely different from noncloned mice (24). Previous study on Dolly, the first clone sheep in the world, reveals that it suffered from serious arthritis, which may serve as a direct evidence of the risk of evading the natural selection on genomic information (1). Together with the identification of reprograming recalcitrant genes in human iPS cells indicate that there is no natural basis to speculate that artificial reprograming can generate healthy pluripotent stem cells.
When we look back to the iPS cells, it is amazing to find that reprograming may violate the rules of the natural selection. The process of inducing iPS cell makes cells that cannot become germ stem cells or other pluripotent stem cells under the normal conditions, obtain the capacity to transmit its genomic information to offspring (25). Given that we cannot declare that the iPS cells are necessarily inferior to those stem cells produced through natural selection, the quality of iPS cells is definitely open to doubt. Although the low efficiency of reprograming can be seen as a threshold or screening mechanism (16), undoubtedly, the outcome will not be the same as the natural selection for different standards they base on. Human ES cells also lack the part of natural selection, which should be thoroughly considered before a hasty clinical application.
TUMORGENESIS AND RISK OF INTEGRATION
Another obstacle to the clinical application of iPS cell is the risk of tumorgenesis. Many iPS cell-derived animals develop tumors because of the reactivation of the c-Myc virus (10), chimeric mice derived from c-Myc–free iPS cells showed substantially less tumor formation (9). However, recent study reveals that use of the c-Myc retrovirus did not affect the teratoma-forming propensity of iPS cells (26). Thus, the mechanisms underlying the different teratoma-forming propensities of iPS cells remain elusive, and the risk of tumorgenesis still hinders iPS cell from the clinical use.
Strictly speaking, all genes which are used to induce iPS cell can be classified as oncogenes (14). We can find evidence that Oct4 overexpressed in human breast stem cells (27) and bladder cancer (28). Moreover, the expression of Sox2 can be detected in human gastric carcinoma (29), stomach adenocarcinomas (30), pancreatic carcinoma (31), vater adenocarcinoma (32), malignant glioma (33), breast cancer (34), and brain tumors (35). Nanog has also been demonstrated that the forced expression in hematopoietic stem cells can give rise to T cell disorder (36). Klf4 is overexpressed in squamous cell carcinoma (37,38), and Klf4 seems to be involved in other cancer formation process.
The nature of these genes may explain why simply excluding c-Myc accomplishes nothing toward a tumor-free iPS cell. Recently, several groups demonstrate that knocking out p53 can enhance the efficiency of reprograming to 10% (39). This discovery is a great achievement in producing iPS cells in terms of efficiency. However, it can be seen from another angle. The fact that p53, a well-known tumor suppressor, can enhance reprograming efficiency implicates the similarities between reprograming process and cancer transformation (40). (Fig. 1).
Moreover, the tumorgenesis is not the sole problem confronting iPS cell. The process of producing iPS cells initially required vectors that integrate into the genome, which can create mutations and limit the utility of the cells in both research and clinical applications. This integration is also seen as a possible cause of tumorgenesis due to the exogenous DNA brought in. Reports demonstrate that the mutation caused by integration in iPS cells is permanent, and can possibly be transmitted to the next generation. Consequently, the application of iPS cell to clinical practice may influence the patients' offspring (8,11).
Such approaches as using nonintegrating adenoviruses to deliver reprograming genes, transient transfection of reprograming plasmids, a certain piggyBac transposition system, Cre-excisable viruses have been reported to diminish the risk of mutation caused by integration (22). Recently, James Thomson's lab reported that they established iPS cell lines through a oriP/EBNA1-based episomal expression system. Because oriP/EBNA1 episomal vectors are gradually lost from proliferating cells in the absence of selection, they get iPS cells completely free of vector and transgene sequences, which are similar to human ES cells in proliferative and developmental potential. Ding and coworkers initially demonstrate that somatic cells can be fully reprogramed into pluripotent stem cells by direct delivery of recombinant reprograming proteins. They also demonstrate that the iPS cells they produce can maintain the state of self-renewal and pluripotency in vivo and in vitro (22). Shortly after this discovery, their demonstration was reinforced by the successful replication of this technology in human fibroblasts (41).
By successfully using recombinant proteins, the possibility of mutation caused by integration of exogenous DNA seems to be wiped out (22). And, this can be seen as a safer approach to generating iPS cells. Yet, it may be not completely safe as it seems to be. There is doubt among some scientists that in terms of the Central Dogma, the genetic risk of exogenous genes may well be realized with their transcription into some mRNAs and then translation into some proteins. Thus, merely avoiding the upstream oncogenes, while using the downstream oncoproteins, would not logically eliminate the risk on the end point of the same line of genetic-chemical flow.
A LIMITATION OF DISEASE MODEL USING HUMAN IPS CELLS
Many labs seek to establish disease-specific iPS cells from patients with a variety of genetic diseases with either Mendelian or complex inheritance for disease etiology. These diseases include Shwachman-Bodian-Diamond syndrome, Gaucher disease Type III, Duchenne and Becker muscular dystrophy, Parkinson disease, Huntington disease, juvenile-onset, type 1 diabetes mellitus, Down syndrome (DS)/trisomy 21, and the carrier state of Lesch-Nyhan syndrome (42), deficiency-related severe combined immunodeficiency (ADA-SCID), as well as β-thalassemia ibroblast cells (25). However, the presence of reprograming recalcitrant genes in human iPS cells will limit these applications. For example, the phenotype difference observed in the patient specific human iPS cells may caused by the genetic background of patients as well as the artificial genetic and epigenetic aberration introduce in the process of iPS cell methods. It is important to consider the problem of the lack of natural selection in human iPS cells in the application of disease etiology.
A CONTROVERSIAL CELL THERAPY
Ever since the Dolly came out, the cloning techniques are criticized for ethical consideration. Because of the involvement of embryos which are generally seen as life, the research on human embryonic stem cells faces great criticism and objection. Thus, the clinical use of ES cells is far from feasible. Many countries have set bans on research in this field. iPS cell, a technique that induce somatic cells into the state of pluripotency, can avoid the use of embryos. Therefore, iPS cell is widely accepted by scientists for circumventing the ethical problems (3,12).
However, iPS cell cannot completely solve or circumvent the ethical problems. The National Institutes of Health Guidelines for Research Using Human Stem Cells forbids research in which hESCs or human induced pluripotent stem cells are introduced into nonhuman primate blastocysts, and research involving the breeding of animals where the introduction of hESCs or human induced pluripotent stem cells may contribute to the germ line. Such rules indicate that iPS cell may not completely circumvent ethic problems.
As it is claimed by many researchers, the iPS cells have an advantage of evading the transplantation rejection, thus emerge as a promising approach to make regenerative medicines (3,12). The patient-specific treatment on the basis of iPS cell seems to resolve the problem of transplantation rejection, which troubles the method of embryonic stem cells. Immunological rejection during the process of transplantation is always a technical problem: patients have to tolerate the side effects brought about by taking immunosuppressive agent to suppress the immune system. However, the tissues or organs transplanted will face the high risk of atrophy and necrosis, which can do great harm to the patient.
However, the advantage that iPS cell can circumvent immune rejection is cast into question. Primarily, it is quite strange to see different labs to use nude mice as the recipients when they try to demonstrate the capacity of differentiation in vivo (4,12). Apparently, it is quite inconsistent that a study aimed at showing a circumvention of the immune rejection barrier by using iPS cell-based therapy would still use sub-lethally irradiated mice as the recipients. If the iPS cells can indeed evade the attack of the immune system, there is no need in using SCID mice to conduct this experiment. Also, many research groups can successfully confirm such expected advantage of iPS cells by injecting iPS clones back to the mice. In theory, this experiment is not so difficult that prevents us from confirming that iPS cell can really circumvent immune rejection, however, we can see little evidence about such experiments. In contrast, evidence in the opposite side accumulates to undermine the claim that iPS cell can successfully evade the attack of immune system in recipients. Dressel et al. (43) demonstrate that “Pluripotent stem cells, including maGSCs, ESCs, and iPS cells can become targets for CTLs, even if the expression level of MHC Class I molecules is below the detection limit of flow cytometry. Thus they are not protected against CTL-mediated cytotoxicity. Therefore, pluripotent cells might be rejected after transplantation by this mechanism if specific antigens are presented and if specific activated CTLs are present.” This discovery indicates that the safety of iPS cells in the clinical use should be considered thoroughly. This issue is not specific to human iPS cells, and this is a general issue underlining in allogenic cell therapy.
Besides, the concept of autologous patient specific iPS cell strategy is flawed if we consider these facts: it is not realistic to repair the disease causing genetic mutations and epimutations in iPS cells and replace with the somatic cells in patients; also impossible to treat the acute diseases and the large tissue degenerations, which usually observed such diseases as cardiac infarction and liver failure.
Definitely, the research on iPS cell is of great value and will remain to be the hot spot in the field of stem cell. However, the focus will be shifted from developing new technique of producing iPS cells to research on the mechanism. Up to now, the process of producing iPS cells remains a mystery. It is essential to know how it comes before hastily applying it to the clinical practice. Also, a promising cell therapy ought to be first based on a correct understanding of life including normal processes and abnormal diseases, and be applicable in vivo.
The authors deeply appreciate Dr Liu SV for providing insightful ideas on iPS cell in communication with us; and Yi Wang, from Washington University School of Medicine in St. Louis, USA, for critical reading of the manuscript, and the anonymous reviewers for their insightful suggestions.