Researchers have successfully reprogrammed human nerve cells back to an embryo-like state by using just a single gene.
This is an inching closer to curing diseases like Parkinson's using cells generated from a patient's own body.
It is known that embryonic stem cells are pluripotent - they can develop into any of the body's cell types.
But such cells are not available in large numbers, as they can only be harvested from a donated egg or embryo, and, for ethical reasons, most countries have laws restricting their use.
In 2006, Shinya Yamanaka and his colleagues at Kyoto University in Japan successfully made mouse cells pluripotent by reprogramming skin cells into a state like embryo cells.
They did so by using retroviruses to insert four genes - known as "factors" - into the cells' DNA.
They repeated the trick a year later with human cells.
However, using genes and retroviruses in this way increases the risk of the cell becoming cancerous, not just because tinkering with DNA has that effect, but also because two of the four factors are known to cause cancer.
In a bid to make these promising cells in a safe way, Hans Scholer's team at the Max Planck Institute for Molecular Biomedicine in Münster, Germany, has been working to achieve pluripotency using fewer factors.
Last year, they did this with the two factors that do not cause cancer, and now they have simplified the recipe further, doing it with just one.
"Remarkably, it turns out that three of these four essential factors are already expressed in human neural stem cells - although not in skin cells - so we only needed to add one factor, OCT4," New Scientist quoted Boris Greber, a member of the team, as saying.
He said that the cells from neural tissue are much easier to reprogram than skin cells, and are less prone to mutations.
It is much harder to get a sample of neural stem cells than skin cells, as it can be done via extracting the cells from the dental pulp of teeth, said Greber.
Inserting even one gene into the chromosome of a cell still permanently modifies its DNA, which is why the new method will remain a lab tool instead of being allowed in the clinic.
However, the researchers are hoping that it will help them improve methods for producing embryonic stem cells.
"Ideally, we will be able to find a chemical that does the same job of expressing the factor without the need for a gene," said Greber.
Earlier this year, researchers in California managed just that when they reprogrammed mouse fibroblasts using a cocktail of proteins.
That technique did not involve inserting genes, and, thus, shouldn't raise the cancer risk. But that was far less efficient.
"Without stable intervention using viruses, the frequency of reprogramming goes down and you have to wait a long time. We don't have the perfect method yet," said Greber.
The study has been published in the journal Nature.