By Elie Dolgin

Researchers have transformed human skin cells into stem cells similar to those in an embryo without using any reprogramming genes, just the viral vector normally used to deliver them.

The findings, reported last week at the International Society for Stem Cell Research annual meeting in San Francisco, Calif., challenge the conventional wisdom about what it takes to produce stem cells that are compatible with a specific patient.

"There needs to be a note of caution and respect for the way the virus works," says Andrew Baker, a gene therapy researcher at the University of Glasgow, UK. "There are a lot of people routinely using the viruses as tools without really understanding them."

Last year, Baker and his colleague Nicole Kane set up their lab to study induced pluripotent stem (iPS) cells, adult cells that have been tweaked to regain their embryonic "stemness." However, before they could get their hands on the genes traditionally used to convert skin cells into iPS cells, they decided to test just the HIV-based lentivirus that introduces the reprogramming factors.

To their surprise, they found that high doses of the virus, encoding a visible marker gene but no reprogramming factors, created cells that looked exactly like pluripotent stem cells. "This was the control experiment that went wrong, effectively," says Kane.

Kane thought that she had contaminated her experiment with iPS cells from a different source. So she sent the samples to an independent forensics lab at the local police department, where a DNA fingerprinting expert confirmed that the iPS cells matched the original skin cells.

Passing the test

Although the efficiency of the reprogramming was less than one in a million and the iPS cells had many chromosomal abnormalities, they passed all the standard tests of pluripotency. The researchers have since repeated their experiment twice, with skin-cell biopsies taken from other people.

"If you push the system a little with a virus, you breach the threshold of toxicity," says Baker. "That makes the cell unstable, and if you provide the right environment, such as adding the necessary culture media, it can induce a cell that satisfies all the criteria for an iPS cell."

By studying where the viruses integrate into the genome, the researchers are now trying to identify other genes that might make reprogramming more efficient. So far, they have flagged several genes involved in cell cycling, microRNA regulation and other processes implicated in the developmental transition that gives the cell stem-cell characteristics, as well as a number of genes known to turn on during cellular reprogramming.

Kane notes that virally derived iPS cells will never be used clinically, owing to concerns about modifying the genome. But the viral vector remains the gold standard for making iPS cells in the lab, so more work is needed to understand how these cells are made. "The findings highlight to the community that we don't have a clue what goes on in reprogramming," says Kane.

But Ernst Wolvetang, an iPS cell researcher at the Australian Institute for Bioengineering and Nanotechnology in Brisbane, was not convinced that the proposed iPS cells were not just skin cells that had become cancerous.

To confirm the cells' true pluripotency, Wolvetang says, the researchers need to repeat the experiment in mice, to see whether virus-only iPS cells introduced into an embryo can produce live animals capable of making sperm or eggs derived from the introduced stem cells--the hallmark test of pluripotency, but an experiment that is ethically impossible in humans. "If you can do this consistently, then I'll buy it," he says.