Monya Baker
Six reprogramming factors in a plasmid reach a holy grail
For the first time, human skin cells have been reprogrammed to pluripotency without requiring genetic elements to insert themselves into the reprogrammed cells. Though so-called induced pluripotent stem cells promise to be as powerful as embryonic stem cells in their ability to differentiate into all cell types, standard techniques use viruses to insert multiple copies of reprogramming genes into the cells; this makes the cells less predictable, and it creates a higher risk of a cancerous growth. As a result, many laboratories have been racing to publish techniques to reprogram cells without permanent genetic modification.
The latest paper, by Junying Yu and James Thomson at the University of Wisconsin–Madison, and colleagues, uses a plasmid that does not integrate into chromosomes to reprogram the cells1. The vector, which is derived from the Epstein-Barr virus but without the viral machinery, is called an episome. It replicates itself only once per cell cycle and is lost from cells over time because copies do not always segregate properly into daughter cells during cell division. Yu and Thomson showed that reprogrammed cells that had not yet lost the plasmid formed teratomas. They then let the cells divide often enough so that the episome was lost. PCR and other tests confirmed that the vector genes were not present, and these plasmid-free subclones expressed the markers, behaviour and morphology of embryonic stem cells.
To get the reprogramming to work, Thomson and Yu tried various ways of arranging the genes on the plasmid. The construct contains all six reported reprogramming factors used in reprogramming human fibroblasts (OCT4, SOX2, NANOG, LIN28, c-Myc and KLF4) plus the SV40 large T gene, which can be used to immortalize cell lines. The reprogramming rates on foreskin fibroblasts are low, only about five colonies per million treated cells, and not all colonies yield stable clones.
Nonintegrating viruses and plasmids have also been shown to work in mouse cells. In the last four weeks, two other techniques to make reprogrammed human cells have been reported, both of which use genetic vectors that first insert themselves into the genome and then excise the reprogramming genes when reprogramming is complete.
"Other vectors were designed so that you can extract the DNA again," says Jeremy Berg, director of the National Institute of General Medical Sciences, which funded Thomson's work. "This never integrates." Not only does non-integration reduce the risk of insertional mutagenesis, it also avoids the non-trivial step of removing genetic vectors after integration.
But Thomson says there is still a lot of work to do. To use the cells for therapies or drug screening, these and all other induced pluripotent stem cells will have to be very carefully assessed to make sure that the reprogramming process itself does not cause mutations or other problems. "You really want a genome that is left in a normal state, and integration is probably a minor risk."
Thomson expects that a wide variety of techniques will be reported in the upcoming months, and then researchers can start to determine which techniques produce the most consistently normal cells.
"The biggest question once the methods are available will be comparing the iPS cells [against each other and against ES cells] and if there are differences, seeing if they matter," says Berg.