July 10, 2007
Turning skin into embryonic stem cells
Introducing four genes into mouse skin fibroblasts reprograms these cells into embryonic stem cells. If similar techniques work in human cells, patient-specific stem cells for tissue engineering and cell-based therapies may be closer to reality

Cut off the limb of a salamander and it grows back completely. Stem-cell scientists and tissue engineers dream of unlocking the same regenerative capacity in adult differentiated mammalian cells. This dream has potentially come closer to reality with three recent reports that describe the ability of four genes to completely reprogram mouse skin cells (fibroblasts) into stem cells possessing many, if not all, characteristics of authentic embryonic stem cells (ESCs)1–3. ESCs are primordial cells that give rise to all of the thousands of cell types in the body, a property termed pluripotency. ESCs can be propagated indefinitely in vitro, genetically manipulated and reinserted into embryos to create animals with tissues and cells derived from donor ESCs. The ability of ESCs to integrate themselves into many different organs suggests the possibility that they may be used to repair damaged or diseased tissues4. As they are the most primitive and versatile cell in the body, making and manipulating ESCs that are genetically identical to the host have become a holy grail of stem-cell scientists. Three routes have been envisioned to make patient-specific ESCs. First, akin to the process used to create Dolly the sheep5, nuclei from adult donor cells can be transferred into egg cytoplasm6 or ESC cytoplasm7, yielding ESCs possessing the donor genotype . The efficiency of this approach is low, and it has not yet succeeded in humans. Second, fusion of adult cells with existing ESCs can reprogram the adult nuclei so that the new cell behaves as an ESC. Unfortunately, the resulting cell possesses two nuclei and thus has four copies of each chromosome instead of two8. These approaches are ethically controversial because they require the donation of eggs or the use of human ESCs. Third, if somatic cells from a patient can be genetically or chemically induced to return to a primordial ESC-like state, then these cells can be directly used as the source to create donorspecific ESCs.

In 2006, Yamanaka and colleagues surprised the cell biology community with their finding that a core set of just four genes reprograms mouse embryonic fibroblasts (MEFs) into cells with ESC characteristics9. Retrovirus-mediated introduction of Pou5f1 (also known as Oct4), Myc (c-Myc), Klf4 and Sox2—all genes known to be involved in maintaining the pluripotency and self-renewal of stem cells—generated “induced pluripotent stem (iPS) cells” that acquired many ESC markers, and in transplantation experiments gave rise to cells in all three germ layers, providing evidence that they were pluripotent. However, genetic analyses showed that the global gene expression patterns and epigenetic modifications, such as DNA and histone methylation patterns that control gene expression, were not identical to those of ESCs. Thus, the extent to which iPS cells could actually replace ESCs in future clinical applications was not clear.

In the new studies, three groups use a different selection process for screening transduced cells and generate iPS cells from MEFs1–3. These newiPS cells possess gene expression programs and epigenetic signatures virtually indistinguishable from authentic ESCs. Intriguingly, the four introduced genes that started the reprogramming process are rapidly silenced in iPS cells as endogenous Pou5f1 and Nanog, two genes that are known to be critical master regulators of ES fate, are activated. The investigators again show that the iPS cells develop into a wide variety of tissues after injection into a blastocyst, and of particular importance, they also show that these cells contribute to the germline to create a new generation of transgenic animals. Jaenisch and colleagues2 go even further and create a mouse comprised entirely of iPS cells, by rescue of tetraploid embryos that can only form placenta.

Several key issues need to be addressed before this technology may be applied to clinical medicine (Fig. 1). First, can a similar process work with human cells? In this vein, continued research with authentic human ESCs will be vital to provide a benchmark of comparison and offer lessons for improvement. Second, the risk of cancer from iPS cells is a great concern Yamanaka and colleagues reported that _20% of all animals derived from iPS cells later develop cancer, coincident with reactivation of the transduced Myc gene1. Previous use of retroviral vectors in humans has caused leukemias. Both Myc and Pou5f1 can promote cancer10,11. A possible solution to this issue would be to transiently deliver the genes. Third, even if patient-specific iPS cells can be produced, methods to direct differentiation and integrate them into the diseased tissue are still needed. Finally, these studies make the point that possibly any somatic cell can be used to clone an individual. Evoking images of Woody Allen’s Sleeper, a nose, a bit of skin, even a drop of blood may be four genes away from being the seed of someone’s clone. Although politicians may seize upon this research to rationalize cutting off funds for ESC research, it is doubtful they appreciate the additional ethical implications if this new approach succeeds.  - Howard Y Chang & George Cotsarelis
NATURE MEDICINE VOLUME 13 | NUMBER 7 | JULY 2007