Differentiated cells can be reset to a state resembling that of embryonic stem cells, but doing so currently requires the use of retroviruses to insert a suite of genes into cell cultures. Because the viruses insert their cargo at random into genomes of individual cells, infected cells derived from the same organism are genetically different, and so it is hard to know whether functional differences between the resulting stem cell lines are due to this genetic variation, to the epigenetic state of the original cells or to chance events.

Now, researchers led by Rudolf Jaenisch at the Whitehead Institute in Cambridge, Massachusetts show a convenient way of generating genetically identical cell populations that can be converted to induced pluripotent stem (iPS) cells by adding a drug1. What's more, they have successfully reprogrammed cells from a number of different organs.

In previous work, Jaenisch's and other laboratories had reprogrammed cells using transgenes that turn on only in the presence of a drug called doxycycline. This allowed them to study how long the transgenes need to stay active for reprogramming to occur. To prove that the cells containing drug-inducible reprogramming genes were indeed pluripotent, researchers then mixed the cells with mouse embryos to create chimaeric mice.

In this latest paper1, Jaenisch and his colleagues show that cells from several different organs in these chimeric mice can be efficiently reprogrammed in vitro by the addition of doxycycline. Cells that have been reprogrammed to give these so-called 'secondary' iPS cells include neural progenitors, mesenchymal stem cells and keratinocytes, as well as cells taken from muscle, intestinal epithelium, the adrenal gland and the haematopoietic lineage.

The secondary iPS cells generated from different tissues from the same transgenic mouse are genetically identical, allowing researchers to examine the effects of cell type and retroviral insertion site on the outcome of reprogramming. For example, the team was able to reprogram intestinal epithelium derived from one secondary iPS cell line, but not from another, suggesting that reprogramming requirements vary between cell types. In particular, the expression of the reprogramming transgenes seems to vary with both cell type and site of insertion in the genome.

The reprogramming rate for the secondary iPS cells was four- to eightfold higher than the production of primary iPS cells, presumably because cells in the transgenic mice already had sufficient proviruses inserted at appropriate sites in the genome. But the overall reprogramming rate is still low, between 2% and 4%.

Jaenisch and his colleagues believe the rate could be because the drug-inducible transgenes may vary in their responsiveness to doxycycline even within genetically identical cells, and also because reprogramming depends on chance events, several of which are required for complete reprogramming to occur.

Nevertheless, a source of genetically identical cells will help researchers home in on these and other variables and greatly simplify the search for methods to eliminate the need for transgenes and create clinically acceptable reprogrammed cells.