a, Nuclear transfer. In this approach, the nucleus of a somatic cell (which is diploid, 2n) is transplanted into an enucleated oocyte. In the environment of the oocyte, the somatic cell nucleus is reprogrammed so that the cells derived from it are pluripotent. From this oocyte, a blastocyst is generated, from which embryonic stem (ES)-cell lines are derived in tissue culture. If development is allowed to proceed to completion, an entire cloned organism is generated. b, Cell fusion. In this approach, two distinct cell types are combined to form a single entity. The resultant fused cells can be heterokaryons or hybrids. If the fused cells proliferate, they will become hybrids, and on division, the nuclei fuse to become 4n (that is, twice the number of chromosomes in a somatic cell) or greater. If the cells are derived from the same species, their karyotype will remain euploid (that is, they will have balanced sets of chromosomes); however, if they are from different species, they will be aneuploid, as chromosomes will be lost and rearranged. Heterokaryons, by contrast, are short-lived and do not divide. As a result, they are multinucleate: the nuclei from the original cells remain intact and distinct, and the influence of one genotype on another can be studied in a stable system in which no chromosomes are lost. If the heterokaryons are of mixed species, the gene products of the two cell types can be distinguished. By altering the nuclear ratio in the fusion, and hence the stoichiometry of the regulators provided by each type of cell, the heterokaryon is reprogrammed towards the desired cell type (Fig. 3). Culture medium also has a role and needs to have a composition favoured by the desired cell type. Dashed arrows indicate slower processes (involving multiple rounds of cell division) than solid arrows (no division). c, Transcription factor transduction. This approach can be used to form induced pluripotent stem (iPS) cells, which have similar properties to ES cells and can be generated from almost any cell type in the body through the introduction of four genes (Oct4, Sox2, Klf4 and c-Myc) by using retroviruses. The pluripotent state is heritably maintained, and vast numbers of cells can be generated, making this approach advantageous for clinical applications. 1n, haploid.

Three approaches to nuclear reprogramming are described: nuclear transfer (blue), cell fusion (pink) and transcription-factor transduction (green). These complementary approaches have provided synergistic insights for almost 50 years and continue to inform the understanding of nuclear reprogramming and influence medical advances. EG cell, embryonic germ cell.

Cell fusion leads to nuclear reprogramming towards a specific phenotype, which is dictated by the nuclear ratio of the fused cell types in heterokaryons, which do not divide. When, for example, cells from humans and mice are fused in a skewed ratio (such as 1:3) (a), the human cells will generally be reprogrammed towards the mouse cell phenotype (three examples are shown). To uncover which genes are involved in this process at the onset of reprogramming (b), genome-wide species-specific gene expression profiling can be carried out on the three types of heterokaryon shown. In this way, the transcripts of human genes that are induced soon after fusion can be identified, and the effects of knocking down these candidate genes (loss of function) or overexpressing them (gain of function) these candidate genes can also be tested45. The function of these genes can then be validated by assays that assess whether they are required for nuclear transfer or for generating iPS cells or induced somatic cells. For example, assays can test whether expression of the genes identified in the heterokaryons with an ES-cell or iPS-cell phenotype (a, centre) enhances, or is required for, the generation of iPS cells or for reprogramming by nuclear transfer. The genes identified in the heterokaryons with a somatic cell phenotype (a, top and bottom) can be tested to uncover whether they enhance, or are required for, the conversion of iPS cells or ES cells into a particular somatic cell type or the conversion of one type of somatic cell into another type. Such experiments will increase the understanding of the molecular regulators of nuclear reprogramming and therefore improve the safety and efficacy of cells produced for therapeutic purposes.

To generate iPS cells, fibroblasts (or another type of adult somatic cell) are transduced with retroviruses encoding four pluripotency factors (SOX2, KLF4, c-MYC and OCT4)56,63. Fully reprogrammed iPS cells have similar properties to ES cells. They are competent to form teratomas on injection into mice and are capable of generating progeny. A patient's cells can be used to derive iPS cells, which can then be induced to undergo differentiation into various types of somatic cell, all with the same genetic information as the patient. For example, dopaminergic neurons could be generated from the cells of a patient with Parkinson's disease and then transplanted to replace those neurons that have been lost. These differentiated cells can also be used in disease models for studying the molecular basis of a broad range of human diseases that are otherwise difficult to study (for instance, those that affect brain cells) and for screening the efficacy and safety of drug candidates for treating these diseases.

The three approaches to reprogramming somatic cells differ in their technical difficulty, speed of reprogramming, efficiency of inducing pluripotency, and cell yield. Therefore, each approach is better suited for studies that provide early mechanistic insights (top) or for therapeutic applications (bottom). The greater the intensity of the colour, the more advantageous the technique. For gaining mechanistic insights (top) into the onset of reprogramming, heterokaryons are particularly advantageous, for three main reasons. First, they are quickly reprogrammed to express pluripotency genes (1 to 2 days ). This is also the case for nuclear transfer. By contrast, it takes weeks to generate iPS cells. Second, reprogramming by cell fusion is highly efficient. When mouse ES cells are fused with human fibroblasts, up to 70% of heterokaryons (enriched by fluorescence-activated cell sorting) activate the expression of pluripotency genes within 1 day. It is technically challenging (and therefore inefficient) to carry out nuclear transfer in mice, so it is difficult to use this approach for large-scale molecular analyses. Furthermore, the efficiency of generating iPS cells by transcription-factor transduction is low, about 0.01–0.1%. Third, cell division does not occur in heterokaryons. It also does not occur after nuclear transfer during the time when pluripotency genes are induced, allowing active mechanisms that induce pluripotency gene expression to be studied because this induction is independent of cell division and DNA replication; passive mechanisms may accompany cell division (for example dilution of DNA methyltransferases). By contrast, many rounds of cell division are required to generate iPS cells. For therapeutic applications (bottom), iPS cells are particularly advantageous, for three main reasons. First, diseases can readily be modelled using iPS cells derived from patients, overcoming the ethical issues and problems with immunological rejection that are inherent in obtaining human ES cells for studying disease. Skin fibroblasts can be readily obtained from the skin of an individual with a particular heritable disease, induced to become pluripotent in vitro and then induced to undergo differentiation to become the cell type of interest (for example a specific kind of cardiac cell). The pathways underlying a disease state (that is, gene expression and signalling) can thus be studied in cells that are not easily accessible in living humans. Second, drug screening can be carried out in vitro using these iPS-cell-based disease models to determine whether therapeutic drug candidates ameliorate or correct aberrant pathways. Third, for certain diseases, cell therapy might soon be used to regenerate or replace defective tissues, with the caveat that the tumorigenic potential, which is in part due to viral vector integration, must be overcome. Both nuclear transfer (leading to ES-cell production) and transcription-factor transduction (to produce iPS cells) have a high cell yield, which is important for cell therapy applications.