In a suite of three papers, two separate teams have transformed human skin cells into cells that are virtually indistinguishable from human embryonic stem cells. This could provide a way to generate all kinds of tissues from individual patients. That would make it easier to study diseases with genetic components and, perhaps, to generate replacement tissue that would not be rejected by patients' immune systems.

At Kyoto University, a team led by Shinya Yamanaka published a paper in Cell showing that differentiated human cells could be reprogrammed to an embryo-like state1 using the same formula that he had previously used to transform differentiated mouse cells2,3. First the researchers engineered cultured human skin cells (called fibroblasts) so that additional genes could be inserted more easily. The team added viruses engineered to introduce 4 genes (OCT3/4 SOX2, KLF4, and c-MYC) into cultured skin cells collected from adults. After several weeks in culture, the team started to see colonies resembling those formed by human embryonic stem (ES) cells. When these cells, called induced pluripotent stem (iPS) cells, were expanded and tested, they were very similar to ES cells. Cell populations doubled at about the same rate; the chromosome-preserving enzyme telomerase, active in ES cells, was also active in iPS cells. Several pluripotency genes that are silenced in fibroblasts but active in ES cells were active in the iPS cells. Genes from the retrovirus were also silenced, indicating that the transformation happens by shifting gene expression of endogenous genes. The iPS cells could be differentiated to make beating heart muscle and proteins characteristic of neurons. They could also make representatives of the three major cells types (ectoderm, mesoderm, and endoderm) in vivo and in vitro. Gene expression assays showed that 1,267 genes had a five-fold difference in expression between hES and iPS cells. More than five thousand genes showed a five-fold difference in expression between fibroblasts and the iPS cells derived from them.

At the University of Wisconsin, a team led by James Thomson and Junying Yu published the success of a similar approach in Science4 Starting with a panel of 14 genes known to be more highly expressed in embryonic stem cells than in other cell types, they also found a quartet of genes that could transform differentiated human cells into a versatile state. In place of KLF4 and c-MYC, this set included the genes for NANOG and LIN28. The researchers used fibroblasts cultured from skin cells from a newborn boy and a fetus. From 600,000 neonatal cells, the researchers generated 57 colonies similar to those of ES cells. These were selected and cultured into 4 cell lines, which behave like ES cells in culture. (Though the work has not yet been published, they say they have also been able to transform cells taken from adults.) These cells expressed genes and surface proteins characteristic of ES cells and, like cells from Yamanaka's team, were capable of forming teratomas, or tumors with cells representing the three embryonic germ layers.

Just after the Cell and Science papers came out, Yamanaka published an additional paper in Nature Biotechnology showing that both human and mouse adult fibroblasts could be reprogrammed using only three genes.5 The transformation took about a week longer to happen and was less efficient but more specific. This finding is significant because the gene that was eliminated, c-MYC, can cause cancer. Earlier studies using c-MYC showed that the iPS cells could be mixed with early mouse embryos and contribute to all types of tissues in live-born chimeric mice, but that these mice were prone to tumors.3 The mouse cells transformed with three genes could also contribute to many of the specialized tissues in chimeric mice, and showed many of the cell-surface markers and other qualities characteristic of high-quality iPS cells. Moreover, Yamanaka also showed that the three genes originally used in reprogramming could be replaced with closely related genes.

Thomson believes they would be very difficult to distinguish from ES cells in terms of behavior, but iPS cells contain several copies of inserted genes. Both Thomson and Yamanaka believe their cells should be studied alongside embryonic stem cells to see what differences exist between them and how much such differences might matter. However, the fact that multiple routes can reprogram differentiated cells increases the likelihood that this technique can be used to create cells for routine drug screening and, perhaps, for cell therapies.