|Home | About | Journals | Submit | Contact Us | Français|
Embryonic stem cells (ESCs) can proliferate indefinitely, maintain an undifferentiated pluripotent state and differentiate into any cell type. Differentiation of ESCs into various specific cell-types may be able to cure or alleviate the symptoms of various degenerative diseases. Unresolved issues regarding maintaining function, possible apoptosis and tumor formation in vivo mean a prudent approach should be taken towards advancing ESCs into human clinical trials. Rhesus macaques provide the ideal model organism for testing the feasibility, efficacy and safety of ESC based therapies and significant numbers of primate ESC lines are now available. In this review, we will summarize progress in evaluating the genetic and epigenetic integrity of primate ESCs, examine their current use in pre-clinical trials and discuss the potential of producing ESC-derived cell populations that are genetically identical (isogenic) to the host by somatic cell nuclear transfer.
Embryonic stem cells (ESCs) are unique pluripotent cells obtained from pre-implantation blastocyst-stage embryos. They can undergo asymmetric division whereby they either duplicate themselves or differentiate into another cell-type. While adult stem cells are multipotent and can only differentiate into a limited number of cell-types, ESCs are pluripotent and are therefore theoretically capable of differentiating into any cell-type. ESCs can proliferate indefinitely in an undifferentiated state . They express specific markers or characteristics including stage specific embryonic antigens, enzymatic activities such as alkaline phosphatase and telomerase, and “stemness” genes that are rapidly down-regulated upon differentiation, including Oct4 and Nanog. Alternatively, they can differentiate in vivo in teratomas into cells representing the three major germ layers: endoderm, mesoderm or ectoderm or they can be directed to differentiate in vitro into any of the 200+ cell types present in the adult body. Since many human diseases result from defects in a single cell type, the potential to replace defective cells by cell or tissue replacement therapy involving differentiated human ESCs (hESCs) provides a possible cure for, or at least the alleviation of symptoms of, various degenerative diseases. Major efforts are currently focused on differentiating hESCs into pancreatic beta cells for use in the treatment of diabetes (Fig. 1) or into the production of substantia nigral dopaminergic neuronal phenotypes with potential applications in the treatment of Parkinson’s disease. In this review, we will summarize progress in evaluating the genetic and epigenetic integrity of primate ESCs, examine their current use in pre-clinical trials and discuss the potential of producing ESC-derived cell populations that are genetically identical (isogenic) to the host by somatic cell nuclear transfer.
As mentioned above, primate ESCs maintain exceptional proliferate capacity in vitro, resembling a property of immortalized or transformed cells while retaining their pluripotent state. Furthermore, in contrast to somatic cells, experiments with murine ESCs have suggested that ESCs exhibit distinctive cancer cell cycle properties allocating more than a half of their entire cycle to S phase with very short gap phases (G1 and G2) . Similar to transformed cells, the cell cycle of ESCs is not sensitive to serum starvation, and their proliferation is not restricted by contact inhibition. The extraordinary capacity of ESCs to replicate, so attractive for regenerative medicine, may also have a dark side since uncontrolled self-renewal is characteristic for most cancer cells, a theme reiterated throughout this review.
Another characteristic of ESCs is their apparent ability to maintain a normal karyotype through large passage numbers. Indeed, a feature of primary cell cultures is the development of abnormal karyotypes as they become senescent. However, recent evidence also suggests that human ESCs may give rise to genetically abnormal cells . Specific abnormal karyotypes have been described such as trisomy of chromosomes 17q and 12, seen within several widely used human ESC lines including H1, H7 and H14. These changes appear to confer selective advantages to the cells during extended culture (passage 30 and over) and in feeder-free (Matrigel-based) culture systems . The gain of extra chromosome 12 or particular regions of this chromosome has been associated with spontaneous testicular germ cell tumors . Other reports have also emerged describing similar chromosomal abnormalities in hESCs related to specific culture techniques. Mitalipova and coworkers  and Brimble and colleagues  described analogous recurrent accumulation of trisomic cells for chromosomes 12 and 17 in hESC lines BGO1, BGO2 and BGO3 routinely propagated by disaggregation into single cells using collagenase, trypsin or cell dissociation buffer. Note that the same cells cultured by manual passaging as clumps displayed normal karyotype even during extended culture periods reaching passages over 100. Triploid cells, apparently, rapidly predominated within the former culture, leading to a population in which the majority or all of cells expressed the abnormality. Interestingly, karyotypically abnormal cells retained normal ESC morphology, expressed canonical ESC markers and maintained ability to differentiate in vitro. On other hand, the abnormal karyotype was associated with significant changes in gene expression patterns, particularly increases in candidate stemness genes. In general, expression of more than 70% of the analyzed genes including Oct4, SOX2, LEFTY2, GABRB3, GBX2 and FGF13 was significantly different in the cells with abnormal karyotype . In contrast, others have reported maintenance of stable karyotypes during extended culture of hESCs [7, 8]. In these cases, sporadic chromosomal abnormalities were present in approximately 20% of the cell population and evidently were not associated with increased rates of cell proliferation or other selective advantages.
Cytogenetic analysis by G-banding of 18 rhesus monkey ESC lines (ORMES series) established and routinely maintained in our laboratory revealed that 15 were karyotypically normal with a diploid set of 42 chromosomes . However, ORMES-1, -2 and -5 displayed various chromosomal abnormalities in the form of balanced 11;16, 5;19 and 1;18 translocations and, in one case, a pericentric inversion within chromosome 1. These changes were stable since abnormalities were detected in all cells of the sample and maintained during extended culture. Chromosomal abnormalities in these ESC lines were discovered as early as passage 9 so it is not clear whether changes were generated during initial establishment or were present originally in the starting embryos. However, it is possible that the splitting method used may also have had an impact, given the fact that the first six ORMES cell lines were exposed to collagenase for disaggregation during isolation and subsequent passaging. ORMES-7 through -18 on other hand, were derived and propagated exclusively by manual passaging.
Chromosomal abnormalities in hESCs destined for therapeutic use are clearly of concern, given the fact that, in vivo, some specific karyotype changes mentioned above are often associated with tumorigenesis. The development of robust culture techniques that maintain stable karyotypes, as evidenced by periodic cytogenetic screening, will be an important advancement towards the clinical application of human ESCs. Commonly, genetic abnormalities are examined on the chromosomal level by G-banding or, less frequently, by fluorescence in situ hybridization. However, it is likely that other less detectable subtle genetic mutations such as microdeletions, minor rearrangements, amplifications and point mutations may also exist that would escape detection by conventional cytogenetic approaches. Indeed, recent genomic analysis of nine hESC lines using high resolution read-out assays suggests that a variety of clonal DNA alterations dominate late passage cells . Four of nine lines generated nuclear DNA copy number alterations during extended culture periods as detected by the Affymetrix GeneChip Chromosome Copy Number Analysis Tool. Copy number aberrations identified in late-passage versus early-passage hESCs ranged from amplifications or deletions of large genomic regions, such as amplification of the entire 17q arm, to more discrete changes including 2-Mb amplification encompassing the MYC oncogene . Interestingly, copy number alterations were associated with genomic regions or genes with altered dosage in human cancers, including gain of chromosome 17 or 20, loss of chromosome 18 and amplification of the MYC oncogene. Using mitochondrial resequencing oligonucleotide array, six mitochondrial heteroplasmic sequence alterations were also identified in two of nine late-passage hESC lines . Mitochondrial DNA mutations are frequently associated with human cancers and also can arise as a consequence of aging. Clearly, these findings suggest that hESCs, like other cells, undergo genetic mutations during in vitro culture and development of higher resolution genetic screening methods is warranted to ensure genetic fidelity of hESCs prior to transplantation.
Mammalian development originates from a single cell (zygote) that upon cleavage gives rise to totipotent blastomeres of the early embryo that eventually proliferate and differentiate into the wide variety of cell phenotypes found in the adult body. The complex pattern of gene expression governing development and differentiation is tightly regulated by epigenetic modifications, i.e. modifications of chromatin not involving changes in the DNA sequence. DNA methylation and histone methylation/acetylation are well known examples of epigenetic modifications. In general, DNA methylation is associated with the silencing of gene expression. Epigenetic errors can arise randomly or under the influence of the environment and often result in disease in humans. For example, DNA methylation has become increasingly implicated in cancer, as many cancer cells contain hypermethylated DNA that in turn can lead to the silencing of tumor suppressor genes by promoter methylation.
Genomic imprinting is a form of the epigenetic program that involves modification of a gene or a chromosomal region that results in absolute or preferential, monoallelic-expression of a specific parental allele. Imprinting genes tend to cluster in the genome and two intensively studied imprinting clusters that have been implicated in human disease are located on chromosome 15q11-q13, known as Prader-Willi syndrome/Angelman syndrome (PWS/AS) region , and the Beckwith-Wiedemann syndrome (BWS) region located on 11p15.5 . Imprinting in these regions is controlled in cis by so-called imprinting centers (ICs) that regulate parent-specific expression of target genes bidirectionally over long distances. Mechanisms involved in the control of imprinted gene expression are complex and poorly understood (for review see ). ICs are subject to parent-specific epigenetic modifications including DNA methylation and histone changes recognized by specific factors such as DNA-binding proteins, that in turn, activate downstream effects leading to appropriate mono-allelic gene expression. These epigenetic modifications must be reprogrammed during development, involving first erasure of old epigenetic marks during germ cell development and establishment of new marks in a gender-specific manner. Methylation of CpG dinucleotides within ICs is proposed to be one of the initial mechanisms differentially marking parental chromosomes in gametes. Once established, locus-specific DNA methylation profiles must be stably maintained in future generations of cells.
As noted above, disruption or inappropriate expression of imprinted genes is associated with severe clinical syndromes and carcinogenesis in humans, thus it is important to address concerns over imprinting integrity prior to transplantation trials. Alterations in the allele-specific expression of imprinted genes, particularly H19 have been associated with embryo exposure to sub-optimal culture conditions [14, 15] suggesting that, in this case at least, the epigenetic mark may be particularly susceptible to in vitro manipulations. Finally, in humans, a link between assisted reproductive technologies (ARTs) and an increased incidence of AS and BWS has been reported raising concerns over the stability of imprinting following in vitro fertilization (IVF) procedures [16–18]. In the case of mouse ESCs, a high degree of internal heterogeneity and instability in H19, Igf2 and Igf2r has also been reported [19, 20].
In evaluating the role of epigenetic regulation, allele-specific expression analysis of imprinted genes is generally conducted based on the identification of allelic sequence polymorphisms in the transcribed regions of the studied genes. Such sequence polymorphisms can subsequently be useful for expression analysis of parental alleles. We have been interested in the epigenetic profile of monkey preimplantation embryos and ESCs and have focused on several imprinted genes. As a prerequisite for this research objective, we identified single nucleotide polymorphisms (SNPs) in the monkey genome that are relevant to determining the parent-specific expression and methylation status of NDN, H19, SNRPN and IGF2 . Based on the existence of SNPs, we have shown an aberrant biallelic expression of IGF2 and H19 in several monkey ESC lines while SNRPN and NDN were normally imprinted and expressed from the maternal allele . In contrast, expanded blastocyst stage embryos, from which these ESCs were derived, exhibited normal paternal expression of IGF2 and maternal expression of H19 suggesting that imprinting marks for these genes were established at this stage of development. Thus, it is clear that the relaxed imprinting of IGF2 and H19 in monkey ESCs occurred during establishment or the initial culture of ESCs. We also showed that this pattern of expression was retained in differentiated neuronal lineages derived from these ESCs indicating, perhaps, irreversible loss of imprinting at this locus. Dysregulation of imprinted genes seen in primate ESCs is most likely a result of improper maintenance of allele-specific methylation patterns in ICs. To further investigate molecular mechanisms underlying aberrant expression of imprinted genes in monkey ESCs, we performed a comprehensive methylation analysis of several ORMES cell lines by investigating the region corresponding to the human IC harboring CTCF-6 binding site . Using sensitive genomic bisulfite sequencing analysis we demonstrated aberrant methylation profiles within IGF2/H19 IC . These altered methylation profiles were associated with biallelic expression of both IGF2 and H19 in these ORMES cell lines. Methylation changes seen in this region most frequently included methylation of the maternal CpG sites along with a few cases of demethylation of paternal alleles.
As mentioned above, altered methylation profiles and associated abnormal expression of imprinted genes has been reported in mouse ESCs . Particularly, the Igf2/H19 IC that was biallelically methylated in ESCs and ESC-derived fetuses. This hypermethylation was associated with biallelic repression of H19, however, Igf2 expression was not biallelic as expected but rather maternal . Similar to monkey ESCs, this aberrant imprinting apparently did not affect mouse ESC phenotype or proliferation capacities. However, injection of ESCs into blastocysts resulted in embryonic abnormalities and lethality in chimeric fetuses . In contrast to mouse and monkey, differentiated populations derived from human embryonic germ (EG) cells show normal monoallelic expression of IGF2 and H19 . This expression profile was also associated with normal monoallelic methylation within the corresponding IC. Interestingly, human EG cells were derived from the gonadal ridge of 5–11 week old embryos and the corresponding stage mouse germ cells exhibit complete erasure of imprints and biallelic expression of H19 and Igf2 . Moreover, human testicular and ovarian germ cell tumors were consistently associated with biallelic expression of both H19 and IGF2 [26–28]. Recently, normal imprinting of several genes in heterozygous human ESC lines was also reported, in particular, monoallelic expression of IGF2 and H19 and normal methylation patterns in the corresponding IC . Thus, these results suggest that hESCs maintain stable imprinting at this locus. However, more extensive propagation of one particular human ESC line to passages over 66 resulted in the activation of the previously silent allele and biallelic expression of H19 .
Biallelic expression of IGF2, often referred as a loss of imprinting, is implicated in BWS and a variety of tumors types including bladder and colon cancers [30–32] and osteosarcoma  suggesting that abnormal (approximately two-fold increase) IGF2 dosage may support tumor growth through autocrine or endocrine effects . It is believed that IGF2 is a potent cell survivor factor that stimulates cell proliferation, and overexpression secondary to biallelic activation leads to oncogenic conditions resulting in uncontrolled cell proliferation, overgrowth and malignant transformation. For example, increased Igf2 expression in primary mouse embryonic fibroblasts stimulated proliferation and also resulted in lack of senescence and rapid conversion to malignancy with tumor formation after transplantation into host animals . Specific conditions that resulted in aberrant methylation in monkey ESCs within the IGF2/H19 IC may have also triggered methylation abnormalities in non-imprinted genes. This scenario could include reactivation of oncogenes as a result of promoter hypomethylation or silencing of tumor suppressor genes caused by hypermethylation. Indeed, a recent study found hypermethylation of promoter regions for several genes associated with cancer development in late-passage hESCs . The promoter methylation status of 14 genes known for their abnormal methylation in cancer tissues was analyzed in nine hESC lines by real-time quantitative methylation-specific PCR. Specifically, increased methylation of three genes including RASSF1, TNFRSF10C and PTPN6 was detected in the late-passage cells compared to early passage ESCs. Notably, hypermethylation within the promoter of RASSF1 was observed in seven of nine late-passage hESCs. RASSF1 is a putative tumor suppressor gene and epigenetic silencing of RASSF1 expression by promoter methylation is a feature of many human cancers, while normal cells display an unmethylated promoter .
The significance of epigenetic aberration observed in monkey and human ESCs to cell function and fate after transplantation of progeny into a recipient remains unknown. A strong argument could be lodged that epigenetic regulation would be of little consequence to terminally differentiated cell populations. Nevertheless, cellular overproliferation and tumor formation resulting upon tissue or cell transplantation are potential clinical problems that must be addressed before clinical trials of ESC-based therapy are initiated.
It now seems entirely reasonable to predict that hESC-derived cell populations will be ready for clinical use within the next 5–10 years. At present, the risks of transplantation are unacceptable primarily because of safety issues and the potential for spontaneous or uncontrolled cellular differentiation after transplantation. The exciting news, however, is that preclinical trials in relevant animal models such as non-human primates are presently being used to address this potential problem along with concerns over what cell populations should be transplanted and why, where transplants should be placed and how well transplanted cells will function and for how long. As pointed out by Dawson and colleagues , the possibility exists that cells could mistarget, i.e., migrate to unintended sites and create a new and different set of problems or misdifferentiate and be of no physiological value to the recipient.
One of the challenges before clinical transplantation studies can begin is to better define and characterize the starting populations of ESCs, for heterogeneity in undifferentiated ESC cultures and inter-line variability are recognized entities. While as many as 200 hESC lines may be available worldwide, Federal funding for hESC research in the USA is currently restricted to Presidential Lines of which 21 are under distribution (http://stemcells.nih.gov/research/registry/). Thus, while private and State supported activities in California, New Jersey and Connecticut may alleviate this limitation in the short term, the need for additional ESC lines is commonly recognized . In non-human primates a number of ESC lines are available, including in the rhesus monkey, 8 as a part of the original derivations by Thomson and coworkers [38, 39] and 19 derived by us in Oregon (ORMES 1– 19); 4 in the cynomolgus macaque  and 1 from a cynomolgus monkey parthenote . In the marmoset, Thomson originally described 8  and more recently Sasaki described 11 . However, it is very clear that existing primate ESC lines derived under “original” conditions, i.e., exposed to mouse embryonic fibroblast cocultures and/or media containing fetal calf serum or serum replacers, are not suitable for clinical use because of the potential for contamination by non-primate pathogens or because they might elicit an immunogenic response that could jeopardize the transplant. To illustrate this point, the report of Martin and coworkers is instructive . These investigators assayed hESCs for the presence of a mammalian sialic acid, N-glycolylneuraminic acid, a membrane component that is not normally found in humans. This sialic acid was found not only in hESCs but also in EBs derived from those cells with the major source likely the serum-replacement-containing medium used in routine hESC maintenance. As the authors conclude, the finding that current conditions carry risks of infectious disease transmission and immunologic sequelae necessitates the production of new lines under xeno-free conditions. Recent progress towards this goal has been achieved by Stojkovic and coworkers  wherein hESCs were grown on allogenic feeders (fibroblast-like cells derived from the spontaneous differentiation of hESCs) in the absence of foreign protein or factors. Additionally, Genbacev and coworkers  derived hESCs under serum-free conditions (the medium did contain knockout serum replacement, however) and then maintained undifferentiated cells on pathogen tested, human placental fibroblast feeders. A similar report involving rhesus ESCs has also appeared .
Inter-line variability in primate ESCs was noted above and the possibility also exists for intra-line variations. A wide variety of culture conditions has been used for hESC maintenance and propagation (for review see Hoffman and Carpenter ) and it is likely that ESCs can actually adapt to the culture conditions or that the conditions themselves result in the selection of a unique population of ESCs. This may also be the case upon low temperature storage where cryosurvival rates are very low, a common result before recent technical improvements became available . In any case, intra-line variability may result in terms of differences in growth rates or differentiation potentials. A quantitative method for the evaluation of such variability may require “stemness” profiling and/or subjecting cells to fixed differentiation protocols. Unfortunately, at present, neither of these options is routine although an extensive description of the hESC transcriptome has appeared recently . The differentiation potential of primate ESCs is primarily qualitative, as in evaluating mesodermal differentiation by observing the presence of contracting cardiomyocytes. This could be useful if the relative efficiency of cardiomyocyte production was the measured outcome or if early markers for a given phenotype could be quantitated in progenitor cell populations.
In approaching preclinical trials, practical limitations must be addressed such as the ability to produce large numbers of cells within acceptable time frames. Moreover, production for clinical use must be accomplished using good manufacturing practices not commonly associated with basic research laboratories. Of course, as noted above the issue of uncontrolled differentiation after transplantation must be addressed in animal models. Currently, investigators typically examine populations of ESC derived cells prior to transplantation for the presence of undifferentiated cell markers such as Oct4 and then look for evidence of tumor formation following transplantation. Limitations to this approach are first that the detection of undifferentiated cells (Oct4 positive) must be 100% effective because the presence of a single pluripotent cell could in theory result in a malignant tumor. Secondly, the evaluation of tumor formation post-transplantation has mostly concentrated on the area at and around the injection site and then over only short time frames. The demands of evaluating tumor formation throughout the body in long-lived primates are clearly intimidating.
With regards the ideal cells for transplantation, highly homogenous, terminally differentiated phenotypes come to mind and there is currently a great deal of research attention focused on producing such populations. From a theoretical perspective, however, it may be wise to consider transplantation of cell populations containing functional as well as progenitor cells as described below in the study by Fair and coworkers . The former would provide immediate activity but for a limited time interval as they are terminally differentiated while the latter may be able to complete differentiation at the transplant site under the influence of the local environment. Thus, the progenitor populations may serve as a source for long term solutions, i.e. they can be expanded and can provide the source for cell differentiation into functional phenotypes. Related to this issue is how many cells should be transplanted and where. Numbers are difficult to quantify because of cell clustering but in excess of a million seems to be the existing target. The calculated yields at the transplant site following sacrifice of the recipient vary substantially from a few percent up to 25–30 percent depending on the local environment. Obviously survival will be dependent on the mechanism for transplantation; liver, heart or pancreas by direct injection into the circulation or the liver; Parkinson’s, injection into substantia nigra the site where dopaminergic neurons are depleted. The former are relatively simple approaches that could be repeated easily over time as conditions warrant versus the latter brain injections that might not be easily repeated. Finally there is the issue of how long transplanted cells function and at what level. We shall see two very different outcomes in the preclinical trials discussed below.
Common approaches to the enrichment of specific phenotypes include the induction of differentiation using specific cells in coculture and/or the inclusion of specific growth factors. Differentiated populations can be further purified by cell sorting, (fluorescence activated cell sorting (FACS) for instance) based on the detection of specific cell surface markers or the expression of a reporter gene such as green fluorescence protein (GFP). An approach that combines genetic manipulation and cell sorting involves the incorporation of a gene construct that restricts reporter gene expression to a specific cell type. As an example, we have generated a gene construct comprised of the human insulin promoter (HIP) coupled to eGFP . This construct when transfected into ESC derivatives supports GFP expression in only those cells that are expressing HIP. Clearly, this provides a unique basis for cell identification and subsequent purification by FACS. In our preliminary, proof of principal studies, we grew rhesus monkey ESCs into EBs and through an expansion stage before plating as a monolayer on a glass coverslip. Cells were transfected with the HIP:GFP construct using ExGen 500 and allowed to differentiate in expansion medium for 3 weeks followed by 2 weeks in differentiation media containing 1nM Exendin 4 as the inducer. When cells were then fixed and co-stained with DAPI (a DNA stain) and insulin or c-peptide, dual labeling was observed indicative of nucleated, insulin producing cells. This experiment will now be repeated and live cells separated by FACS will be characterized for homogeneity and ultimately used in transplantation studies.
Here we will review two preclinical trials involving the production, isolation and transplantation of ESC-derived phenotypes. The first example illustrates the advantages of the mouse model and relates to the relief of hepatic dysfunction with ESC derived endodermal precursors . Factor IX, produced by the liver, is a critical component of the clotting system. Factor IX knockout mice are available and they survive for only 7 days without intervention. These investigators injected, directly into the parenchyma, 1 million eGFP expressing precursor cells and studied Factor IX activity in the circulation over time. Injected animals lived for up to 115 days and displayed Factor IX levels at about 10% of control. Upon termination and characterization of liver sections, no evidence for neoplasia was found and GFP mRNA was detectable throughout the livers by PCR. Co-localization of Factor IX and GFP supported the conclusion that Factor IX was coming from the transplanted GFP-positive cells. Thus the authors conclude that functional, persistent hepatic engraftment of putative endodermal precursors occurred and that they obtained proof of principle that directly injected differentiated ESCs are likely to have clinical relevance for replacement of hepatic function.
The second example involves the nonhuman primate, a clinical relevant animal model that is especially appropriate for neurodegenerative disease processes, in this specific case, the recovery and transplantation of ESC-derived dopaminergic (DA) neurons for the treatment of Parkinson’s disease . These studies also provide an example of induced or preferential differentiation into a desirable phenotype, a process undoubtedly involving both selection or induction of some phenotypes and the deselection or suppression of others. In a 2-step protocol, monkey ESCs were first removed from mouse embryonic fibroblast coculture and grown on PA6 mouse stromal feeder cells. This step resulted in ESC exposure to stromal cell-derived inducing activity, shown previously to enrich neural progenitor cell populations. Parenthetically, methods for determining success or failure of such an approach include immunocytochemistry (ICC) for specific markers, for instance; the glial fibrillary acidic protein, GFAP, for astroglial, galactocerebroside C for oligodendroglial and TuJ1 and Map2ab for neuronal cells. Finally the resulting neurospheres were exposed to the growth factors, BGF2 and BGF20. Under this differentiation protocol, which required 21 days, 24% of the TuJ1 positive population (neuronal cells) was also tyrosine hydroxylase (TH) positive (dopaminergic neuronal phenotypes). Expression of mesencephalic DA neuron markers were confirmed by reverse transcription polymerase chain reaction (RT-PCR) and these cells released dopamine in response to high K+ depolarizing stimuli. When 300,000 to 600,000 cells were injected into each side of the bilateral putamen of chemically lesioned monkeys, modest improvements in neurological scores were observed over sham operated controls (behavior) and in 18F-fluorodopa uptake and immunohistochemical imaging. As stated by the authors, these studies may represent the first demonstration of the efficacy of transplantation therapy using ESC derived DA neurons in an experimental primate model of Parkinson’s disease.
Finally, it should be noted that a significant proportion of the pre-clinical ESC-transplantation research is currently focused on investigating how ESC-derived cardiomyocytes will respond inside the heart in vivo. Kehat et al demonstrated that human ESC-derived cardiomyocytes when transplanted into pig hearts could act as biological pacemakers , Laflamme et al demonstrated that human ESCs could form human myocardium in the rat heart  and Menard et al demonstrated that murine ESC-derived cardiomyocytes could significantly regenerate infarcted sheep hearts .
Despite the remarkable strides that have been made to date with allogenic ESCs, these cells are genetically divergent from the host/patient and would therefore inevitably undergo immunological rejection over time. Even using an extensive daily cocktail of immunosuppressive drugs, the majority of heart transplant patients will undergo at least one episode of graft rejection within their first year, requiring increased amounts of immunosuppressive drugs (increasing the patient’s risk of infection or cancer) and 40% of heart transplant patients will die within ten years . While we do not know how severe the intensity of graft rejection would be for ESC-derived cells injected into human heart muscle, the only sure way to avoid all immune rejection in the long term would be to transfer genetically identical (isogenic) differentiated ESCs back into the patient, and the only method currently known for producing isogenic karyotypically normal pluripotent embryonic stem cells is somatic cell nuclear transfer or “cloning”.
The concept of cloned pluripotent embryonic stem cells resulted from the synthesis of two separate experimental fields: somatic cell nuclear transfer and embryonic stem cell research. Somatic cell nuclear transfer, or cloning, dates back to 1962 when Dr. John Gurdon first demonstrated that a differentiated vertebrate somatic cell nucleus (from a larval stage intestinal epithelial cell) could reprogram back into an embryonic state after being transferred into an enucleated Xenopus laevis egg and elicit the development of a cloned adult frog . Following this groundbreaking result several decades of amphibian and mammalian nuclear transfer research were conducted culminating in the announcement in 1997 of the first mammal to be cloned from an adult donor cell nucleus . To date sheep , mice , cows , goats , pigs , rabbits , cats , horses , rats , and dogs  have all been cloned from differentiated adult somatic donor cells, even Rhesus monkeys have been cloned, albeit from embryonic donor cells [69, 70]. Embryonic stem cells – or at least ES-like cells – have been derived from embryos of multiple species, including: mice , sheep , hamsters , cattle , pigs , mink , rabbits , rats  and non-human primates . In 1998 the first stem cell lines to be derived from human embryos were obtained . The conceptual unification of these two separate scientific fields (nuclear transfer and embryonic stem cell research) suggested that it might be possible to produce pre-implantation human embryos by somatic cell nuclear transfer and then derive isogenic embryonic stem cells from the resulting cloned embryos [79, 80]. Human nuclear transfer embryonic stem cells (ntESCs) obtained in this way could be used in both “research cloning” and “therapeutic cloning”. Research cloning would involve using ntESCs to study specific diseases in vitro (see Section 5.2) and therapeutic cloning would involve differentiating ntESCs into therapeutically useful cells that could be transferred back into a patient suffering from a degenerative disease (see Section 5.3).
As previously mentioned, research cloning involves using ntESCs to study specific diseases in vitro. By comparing healthy stem cells with stem cells cloned from patients with specific genetic diseases scientists may be able to gain insights into certain illnesses, study genetic diseases in which the genes involved have not been identified and test potential medicines in vitro. As well as identifying drug targets and testing potential therapeutics, ntESCs could be used for patient specific toxicity testing and studying differentiation pathways. One of the main proponents of research cloning is Dr. Ian Wilmut who has recently been granted a UK Human Fertilisation and Embryology Authority license to derive human ntESCs to study motor neuron disease (MND). MND is caused by the death of motor neurons in the brain and spinal cord. This progressive neural degeneration causes severe disability from the outset and usually results in death within three to five years. Twenty years of genetic research has identified only one gene (SOD1) which accounts for only 3–5% of all MND cases. Dr. Wilmut’s team plan to generate ntESCs from MND patients. By turning these MND-defective ntESCs into motor neurons they will have a unique opportunity to identify the disease-causing genes and discover what causes these cells to degenerate. The cultured ntESC-derived motor neurons could also be used to discover drugs that effectively block or reverse the progression of MND. Other suggested examples of research cloning include deriving ntESCs from patients with early onset Alzheimer’s disease or autism and then performing mechanistic studies after differentiating these ntESCs into neuroprogenitors . Assuming that ntES cells maintain a relatively stable epigenetic state, then it seems reasonable to suggest that research cloning may soon provide a powerful technique for gaining insights into certain diseases, identifying disease-causing genes and testing potential drugs/treatments in vitro for a wide range of human genetic afflictions. However, as a caveat, it is important to point out that, to date, the usefulness of these cells, even for disease models, remains unproven.
Differentiated human embryonic stem cells express major histocompatibility complex I (MHC-I) antigens . This means that transplantation of genetically unrelated differentiated ESCs into a patient (without immunosuppressive drugs) will incite an immune response and result in the rejection of the transplanted ESCs. Strategies to avoid rejection include: banking ESCs with defined MHC backgrounds, genetically manipulating ESCs, transferring to immune-privileged sites and therapeutic cloning. Banking ESCs with defined MHC backgrounds is unlikely to solve the rejection problem as the number of allele combinations for the MHC-I genes alone is more than one billion diploid combinations, making the probability of a perfect match extremely improbable . Genetically manipulating ESCs to remove the MHC rejection is also problematic, one suggestion would be to replace via homologous recombination every single MHC allele in every cell of an ESC line with each patient-specific MHC allele, and this long and technically difficult procedure be repeated for every single candidate patient, a somewhat unfeasible proposition. The other genetic manipulation suggestion would be to remove the MHC epitopes altogether to make a “universal” ESC line. This universal ESC line proposition is significantly damaged by the fact that elimination or even down-regulation of MHC epitopes initiates a cytotoxic response by recipient NK lymphocytes, caused by MHC-I proteins’ inhibitory effect on natural killer (NK) cells . Transferring unmatched differentiated ESCs into immune-privileged sites would certainly prolong cell survival, but these sites are restricted to the brain, eye and testis, severely limiting the therapeutic applications. Therefore, the only theoretically feasible way to avoid the MHC-I immune rejection of differentiated ESCs is to generate cloned human ESCs that possess identical MHC-I molecules as the patient. The concept of regenerative cell transplantation based on the establishment of isogenic patient-specific human embryonic stem cells derived from blastocysts cloned from a patient’s own somatic cell nuclei is commonly referred to as “therapeutic cloning”.
The first successful derivation of human ESCs from cloned blastocysts was in 2004 . Dr. Hwang’s group optimized the nuclear transfer protocol to obtain a cloned human blastocyst development rate of 29% . Specifically, Dr. Hwang allowed the fused donor cell nucleus to reprogram inside the enucleated oocyte for two hours, activated with 10µM ionophore/DMAP, cultured the cloned embryos for 48 hours in G1.2 media and then transferred them into hmSOFaa culture media for the rest of their in vitro development . In addition to these optimizations, Dr. Hwang’s group also only used very fresh oocytes donated from fertile women and then barely allowed them to mature before enucleating them using a very gentle squeezing technique that may cause less damage than typical aspiration . Using this improved protocol Dr. Hwang was able to obtain significant quantities of cloned blastocysts from which he managed to derive a human ntESC line. However, the overall efficiency of deriving this ntESC line was very low with 242 human oocytes being used to derive one ntESC line . In order to improve the ntESC derivation efficiency Dr. Hwang made several other optimizations to his experimental protocol. These included using human feeder cells, limiting the trypsin exposure of the donor cells to 30 seconds (while monitoring for cellular damage), keeping the oocyte hyaluronidase exposure to a minimum, deriving ESCs directly from the nuclear transfer blastocysts rather than using immunosurgery and performing large amounts of practice nuclear transfer (typically using non-human oocytes) to improve his technicians’ micromanipulation skills . Using this superior protocol Hwang was able to increase the ntESC derivation efficiency to the extent that he could get, on average, one ntESC line from every 17 human oocytes used . When oocytes only donated from women under 30 were considered, Dr. Hwang was able to get one ntESC line for every 12 human oocytes used. With an average superovulation producing 10 to 20 human oocytes  this new improved ntESC derivation efficiency significantly increases the feasibility of therapeutic cloning. Dr. Hwang also demonstrated that each human ntESC line he derived was pluripotent, chromosomally normal and expressed the exact immunologically identical MHC antigens as the donor patients . This suggested that human ntESCs could be differentiated into multiple therapeutically useful cell-types, that these cells were karyotypically normal and that these cells would not be rejected following transfer back into the donor patients. Despite the phenomenal strides that have been made in the therapeutic cloning field, several issues remain to be resolved. First, Humpherys et al have demonstrated that cloned mice demonstrate significant aberrant gene expression, probably due to a failure of the enucleated oocyte to fully reprogram the donor cell nucleus back into an embryonic epigenetic state . Therefore, we should ask if ntESCs demonstrate aberrant gene expression in comparison to IVF ES cells, and if so by how much? It should be noted that the human ntESCs analyzed to date all demonstrated the same morphology and stem cell markers as IVF ESCs  suggesting that any differences would probably be negligible; nevertheless further comparative research in this area is recommended. Second, ntESCs possess the same nuclear genes as the donor but express mitochondrial epitopes from the enucleated oocyte’s mitochondrial DNA. It has been suggested that mitochondrial heteroplasmy could influence ntESC stability and differentiation  and further research in this field would be prudent. However, Lanza et al did not find any rejection of bovine differentiated ntESCs following transplantation – even though they expressed allogenic mitochondrial alleles  – suggesting that mitochondrial heteroplasmy does not have a significant negative effect. In addition to aberrant gene expression and mitochondrial heteroplasmy, other probably minor factors that should nevertheless be investigated further include the lack of sperm-derived centrosomes in ntESCs, the effect of non-random X-inactivation in ntESCs and the possibility that ntESCs may possess the shortened telomeres of their somatic donor source . Assuming that these ntESC-specific issues are demonstrated to have negligible negative impact, we can then ask: is therapeutic cloning a viable medical option?
Dr. Hwang’s research demonstrated that, on average, one human ntESC line could be obtained per superovulation. As several rounds of superovulation are typically required to achieve a pregnancy via human IVF this suggests that therapeutic cloning, at least from an efficiency perspective, is as viable as a medical treatment as IVF is as an infertility treatment. However, when a patient has a specific genetic defect then the ntESCs would also possess the same genetic defect. For example, a patient with a defective insulin gene (type I diabetes) would produce cloned ntESCs with the same defective insulin gene, eliminating the possibility of producing differentiated ntESCs that secreted significant quantities of insulin. In these cases therapeutic cloning would not be enough to produce a medical treatment and the ntESCs would have to undergo a round of gene therapy to correct the genetic defect in question. In our type I diabetes example the ntESCs would need an insertion (via homologous recombination) of the functioning insulin gene before being differentiated and transplanted back into the diabetic patient. The combination of therapeutic cloning with gene therapy has already been successfully performed in the murine model . Rag2−/− mice are not able to produce lymphocytes and therefore are severely immunodeficient. Rideout et al produced murine ntESCs from Rag2−/− immune-deficient mice and then replaced the defective Rag2 allele in the ntESCs with a functional Rag2 allele via homologous recombination. These genetically-corrected ntESCs were then differentiated into hematopoietic precursors and transplanted back into Rag2−/− mice. These genetically-corrected ntESCs were able to differentiate into mature B and T lymphocytes therefore providing a significantly increased level of immune function in the recipient Rag−/− mice . This research provides a paradigm for the treatment of certain genetic diseases using a combination of therapeutic cloning and gene therapy. Murine ntESCs have also been used to cure the murine equivalent of Parkinson’s disease . Barberi et al differentiated murine ntESCs into dopaminergic neurons and transplanted these ntESC-derived dopaminergic neurons into parkinsonian mice. The grafted mice demonstrated a significant alleviation of their parkinsonian phenotype with no aberrant differentiation or teratocarcinoma formation observed . Both the work of Rideout et al and Barberi et al demonstrate that therapeutic cloning (in combination with gene therapy in certain cases) can be used to cure diseases in the murine model. This research provides hope that therapeutic cloning will eventually be used to cure human degenerative diseases using the same basic techniques demonstrated in the mouse [89, 90].
In this review, we have summarized the progress in evaluating the genetic and epigenetic integrity of ESCs, examined their current use in pre-clinical trials and discussed the potential of producing ESC derived cell populations that are genetically identical (isogenic) to the host by somatic cell nuclear transfer. In this section we will discuss some of the future potential applications of ESCs, describe what we believe to be a prudent approach towards achieving these goals and present our overall conclusions.
There are three general areas in which human ESCs could be used: human developmental biology, drug discovery and transplantation medicine. In the field of human developmental biology, ESC study will enable us to investigate in detail how different cell-types are derived. This basic biological information will help us to further unravel the mysteries of human development, laying the foundation for future clinical applications and allowing us to identify the biological effects of chemicals/reagents to observe if they have teratogenic effects. In the field of drug discovery, large numbers of drugs could be quickly screened in cultured ESCs for a fraction of the cost of screening the same number of drugs in human patients. In the field of transplantation medicine, ESCs have the potential to treat a plethora of diseases. They could be differentiated into neurons and glia to treat neurological diseases such as Parkinson’s and Alzheimer’s, muscle cells to treat muscular dystrophies or heart disease and insulin secreting islet cells to treat diabetes (see Fig. 2). Theoretically, ESCs should be able to cure or at least ameliorate the effects of almost any degenerative disease or injury. However, despite the panoply of therapeutic applications for ESCs, there are still a number of issues that need to be resolved before ESC-derived cells are utilized in clinical trials.
In Section 2 we discussed the genetic integrity of ESCs. The majority of primate or human ESCs do maintain a normal karyotype, even after extended culture. However, we would suggest that higher resolution genetic screening methods should be used to investigate whether minor genetic mutations are endemic within the ESCs before human clinical trials are conducted. In Section 3 we discussed the epigenetic stability of ESCs. Primate ESCs show abnormal expression of the imprinted genes H19 and IGF2 (see Section 3) and recent evidence has found that human ES cells, while epigenetically stable at lower passage numbers, can also show aberrant imprinted gene expression (of H19) after extensive propagation . It remains to be seen whether this epigenetic instability will affect the transplanted ESC-derived cell population. In Section 4 we discussed the preclinical trials of ESC-derived cell transplants. ESC-derived cells have been used to restore hepatic function in Factor IX knockout mice and ameliorate the equivalent of Parkinson’s disease in the experimental primate model (see Section 4). However, significantly more animal data of this caliber is required before human clinical trials can be conducted. In Section 5 we discussed nuclear transfer ESCs that are genetically identical (isogenic) to the patient. These isogenic cells have their own specific issues that need to be investigated. Would maternal mitochondrial epitopes induce an immunological response? Might there be any negative effect due to the lack of sperm-derived centrosomes, non-random X-inactivation or possibly shortened telomeres in ntESCs? These are all questions that must be addressed before human clinical trials can be conducted.
As we have discussed, there are many important issues that must be resolved before any ESC-derived cells get transplanted into sick adults or children. With the possibility of these cells becoming cancerous and doing more harm than good we really need to take a prudent approach towards advancing ESCs into human clinical trials. While the vast majority of animal research is conducted in the rodent, we believe that the questions that we have posed in this review would be best addressed using the non-human primate. Rhesus macaques (the standard non-human primate model organism) possess a remarkable anatomical, physiological and metabolic similarity to humans and many human neurological diseases such as Alzheimer’s and Parkinson’s can only be accurately modeled in the non-human primate. In addition to the biological similarities between rhesus macaques and humans, the ESCs derived from rhesus or human blastocysts also demonstrate extensive similarities not observed in murine ESCs. Murine ESCs proliferate quickly, while rhesus and human ESCs proliferate at a lower rate . Murine ESCs express SSEA-1, while rhesus and human ESCs express SSEA-3 and 4. Finally, murine ESCs are induced by LIF to maintain an undifferentiated state, while rhesus and human ESCs are induced by bFGF . In summary, rhesus macaques are better model organisms than mice for human diseases and rhesus ESCs demonstrate extensive similarities to human ESCs that murine ESCs do not. This leads us to conclude that using rhesus macaques will provide significantly more pertinent data following ESC-derived cell transplants and provide an accurate model for developing strategies to prevent immune rejection and test the safety and efficacy of ESC-based medical treatments.
Embryonic stem cells promise to open a new window in human existence. They possess the unique potential to replace our cells as they age, mutate and die. This scientific advance offers us the tantalizing possibility of maintaining our bodies in a state of mental and physical well-being inconceivable even one generation ago. As we enter the new millennium, the gift we are being offered is nothing short of the chance at longer, healthier lives.