A major impediment to clinical application of a technique for creating embryonic-like stem cells without using embryos has been removed, at least in principle.

Specialized cells can be reset to an unspecialized state capable of becoming any cell type in the body. Though induced pluripotent stem (iPS) cells made from individual patients could be incredibly valuable for drug screening or cell therapies, the established process to create the cells requires using retroviruses to insert several extra copies of genes into each cell. This renders the cells less predictable and more prone to forming tumours, and may make them unacceptable for human transplantation.

Two recent papers in Science show that, at least for certain cell types in mice, viral integration is not necessary; one technique does away with viruses altogether. A team led by Konrad Hochedlinger at Harvard Medical School in Boston used a type of virus that does not insert itself into the genome to deliver the genes necessary for converting cells to pluripotency. Hochedlinger's team focused on converting adult liver cells, which have previously been shown to require fewer sites of viral integration and are more easily infected by the adenovirus vector the team was using. This approach generated cells that passed stringent tests of pluripotency1. If transplanted into mouse embryos, the cells go on to produce a wide range of tissue types within newly born mice, including sperm. Interestingly, though, about a quarter of the cell lines had twice the usual number of chromosomes, an abnormality that is not observed with other techniques used to produce iPS cells.

After using adenoviruses to figure out the best way to combine genes, Shinya Yamanaka at Kyoto University in Japan then tried a technique that did not use a virus. They transfected mouse cells cultured from embryonic tissue (mouse embryonic fibroblasts) with plasmids over several days. The team also generated cell lines that expressed markers of pluripotency and formed sperm in chimeric mice, but the researchers started with specialized embryonic cells instead of adult cells2.

Kathrin Plath, at the University of California, Los Angeles has created mouse and human iPS cells. She says the starting cell type might be very important when using non-integrating viral vectors, but she's confident that the plasmid technique Yamanaka used on embryonic fibroblasts can be made to work using fibroblasts derived from skin biopsies from adult mice.

She also is not particularly worried that the yields of iPS cells are so much lower with this method than those seen when using retroviruses. “The efficiencies are quite low, but at the end if you get them, you don't care.”

Besides efficiency, there are hurdles between this work in mice and being able to reprogram human cells from easily accessible tissue samples like a skin biopsy. The rates at which cells convert to pluripotency are far lower than the rates obtained using integrating viruses, and neither team can be absolutely certain that no fragments of the introduced DNA integrated into the cells' genomes. Nonetheless, the work indicates that permanent genetic modification is not necessary for creating iPS cells. Though the cells still need to be more thoroughly compared to ES cells, iPS cells' potential use for therapy and research seems greater than ever.