Teruhiko Wakayama carried out his analysis after questions were raised about the two Nature papers. Credit: Kyodo/AP/Press Association Images

A genetic analysis of stem cells has raised fresh confusion about two recent papers1,2 detailing a simple method to reprogram mature cells into an embryonic state.

The analysis challenges claims by Haruko Obokata, a biologist at the RIKEN Center for Developmental Biology (CDB) in Kobe, Japan, who said that she had used the method to produce two stem-cell lines from a certain strain of mouse, but both lines have been found to come from strains different from the one she claimed. If verified, the findings suggest that contamination, rather than the much-touted reprogramming method, produced the cells.

“Something is grossly wrong,” says Hiromitsu Nakauchi, a stem-cell researcher at the University of Tokyo.

The genetic analysis, done by Teruhiko Wakayama, a cloning specialist at Yamanashi University, Japan, and a co-author of the papers, was conducted on stem-cell lines that Obokata produced to test the method before publication. They were not the cell lines described in the papers. Genetic tests on the published cell lines showed the expected genetic match between the stem cells and the source mice, Wakayama told Nature News.

Critics say British Columbia's grizzly-bear cull not based on science How the $1,000 genome became reality China's moon rover responds but cannot move

Wakayama’s genetic findings come less than a week after the senior corresponding author on one of the papers1 posted a protocol giving a more detailed explanation of the method to help researchers to repeat the experiment. The move, by Charles Vacanti, of Brigham and Women’s Hospital in Boston, Massachusetts, was intended to address the fact that no one outside the team has yet claimed success in reproducing its method, despite more than a dozen independent attempts. A spokesperson for Brigham and Women’s Hospital told Nature News that Vacanti “is not speaking to media at this time”.

Obokata and her colleagues in Japan and at Brigham and Women’s Hospital announced the technology in Nature on 29 January1,2. (Nature News is editorially independent from the research-publications section of Nature.) The two papers described how stressing mature cells by placing them in an acidic solution or putting them under physical pressure could return them to an embryonic-like state in which they achieved pluripotency — the ability to give rise to any of the body’s cells. The team called the phenomenon stimulus-triggered acquisition of pluripotency (STAP). 

But problems in those papers, including the presence of duplicated images and plagiarized passages, triggered an investigation by RIKEN. Previous work by Obokata, including her doctoral thesis, has also been called into question. RIKEN is continuing to investigate the two Nature papers after preliminary inquiries found four “serious errors” in them, including the unattributed use of long passages from another paper. Obokata has resisted requests from CDB director Masatoshi Takeichi to retract the publications, because she maintains that they describe a real phenomenon. She did not respond to e-mail enquiries about Wakayama’s genetic analysis.

Wakayama says that he decided to carry out his analysis of the STAP cell lines in response to the questions raised about them. The tests were “just simple preliminary tests done in [his] own laboratory”, he says, and they have not been verified by others. He has also passed a batch of 20 STAP-generated cell lines to an unnamed independent institution for in-depth investigation. The institution’s report will not be published for a few months, he says.

On 25 March, Wakayama told Nature News that his preliminary analysis showed that the STAP cell lines on which Obokata worked alone were genetically distinct from the ‘129’ strain of mouse said to have been used to derive them. Instead, he says, they came from two other strains, known as B6 and 129B6F1 hybrid. “This discovery was a shock,” says Wakayama.

The two genetically mismatched cell lines were not part of the experiments published in Nature — they were the result of an experiment that Obokata performed in the build-up to publication, after Wakayama gave her a mouse from the 129 line to see whether the STAP method would work with strains of mice other than the one used for the paper (which was the 129B6F1 hybrid). The paper mentioned success with the 129 line in passing in the methods section of the main paper, but did not describe any data from STAP cells derived from it.

Wakayama says that whenever he gave Obokata other lines of mice to test, such as the 129 line, she “would produce STAP cells without fail”. But the data for those other lines were “for some reason not included in the Nature papers”, he says, adding: “This result does not call into question the Nature papers.”

Wakayama also performed a genetic analysis of some of the STAP stem-cell lines for which data were described in the paper. As expected, they showed a genetic match with the 129B6F1 strain that was used in the experiment.

Wakayama says that RIKEN and the independent institute are now discussing his findings. A RIKEN press officer told Nature News: “At present, I cannot confirm the details.”

Contamination of stem-cell lines can happen. “But the chance of having two STAP cell lines contaminated by different embryonic stem lines is very, very small,” says Nakauchi, who has not seen Wakayama’s analysis.

“This is really, really confusing,” says stem-cell scientist Hans Schöler, who works at the Max Planck Institute for Molecular Biomedicine in Münster, Germany, and was not involved in the Nature papers. He has not seen Wakayama’s analysis, but he says: “So far I do not see any proof of misconduct. Moreover, I am still not convinced that STAP cells are bogus.”