A Californian team say they have managed to convert human skin cells directly into functioning brain cells.
The scientists manipulated the process by which DNA is transcribed within foetal skin cells to create cells which behaved like neurons.
The technique had previously been demonstrated in mice, says the report in Nature.
It could be used for neurological research, and might conceivably be used to create brain cells for transplant.
The scientists used genetically modified viruses to introduce four different "transcription factors" into foetal skin cells. These transcription factors play a role in the "reading" of DNA and the encoding of proteins within the cell.
They found the introduction of these four transcription factors had the effect of switching a small portion of the skin cells into cells which functioned like neurons.
Unlike other approaches, the process did not involve the reprogramming of the skin cells into stem cells, but rather the direct transformation of skin cells into neurons.
Marius Wernig, an assistant professor of pathology at Stanford University School of Medicine in California, was one of the researchers.
"We showed that it is possible to convert human skins cells directly into nerve cells which look and behave like nerve cells which usually only exist in the brain," he told BBC News.
"It was known that it was possible to change a specialised cell back into a stem cell, what's called an induced pluripotent stem cell (iPS), but it was not known whether a specialised cell could be pushed into another direction, other than backwards."
Professor Wernig conceded that there were examples, some dating back many years, where specialised cells have been switched into similar cell types, but he believes this is the first example of where cells have undergone such radical conversion.
He believes the immediate application will be in modelling diseases, whereby skin cells from a patient with a known neurological condition could be used to produce new brain cells for research.
"It is very very difficult to look into the brain. There is a big skull which protects the brain very well and therefore it's difficult to image," he said.
"Everything that can be done at a cellular level is only possible after a patient has died, by which time the disease is usually in the final stages and you have no chance of seeing how the disease develops."
The technique might one day also be used to create new brain cells which could be transplanted into patients with neurological disorders, he said.
Created from the patient's own skin, these cells would be an exact match for the patient, although there would be many obstacles to overcome, not least the challenge of producing enough of the right type of brain cells.
Commenting on the study, Jim Huettner, an associate professor at Washington University School of Medicine, said the research was "convincing and important".
"They have shown similar things in mice before but in humans they've discovered some subtle differences which often turn up when moving from mice to humans," he said.
"But the work solidifies the idea that this kind of transition is possible and that it's not just some fluke in the mice model."