TOKYO—Since January, scientists around the world have unsuccessfully attempted to reproduce a surprising stem cell finding that claimed simply stressing adult cells could turn them into powerful stem cells that resemble those found in early embryos. Now, ScienceInsider has learned that some of the labs involved in producing the two papers describing the work had not attempted to reproduce the technique before the papers were published. Only two of the labs involved in the papers say they have been able to generate so-called STAP cells.
The controversy emerged about 2 months ago, when stem cell scientist Haruko Obokata suddenly became a media sensation in Japan. The 30-year-old was widely celebrated when she and colleagues published two papers in Nature describing a new and surprisingly simple way of creating stem cells, which the researchers dubbed stimulus-triggered acquisition of pluripotency (STAP) cells.
Now, Obokata is still in the spotlight, but the script has changed. Allegations that the papers contain images recycled from Obokata’s Ph.D. thesis, among other problems, have fed doubts about the claims. And as the scrutiny has grown, several of the collaborating researchers have confirmed that they have not yet produced STAP cells either.
It would not be unusual for the lead author to perform most of the work for the papers, says Martin Pera, a stem cell scientist at University of Melbourne in Australia. But given how revolutionary—and simple—the technique was supposed to be, he says, “the uncertainty concerning reproducibility is puzzling.”
Earlier methods for deriving pluripotent stem cells, the all-purpose cells valued for their possible role in regenerative medicine, required extracting them from embryos or forcing adult cells to overexpress certain genes. But in an article and a letter online in Nature on 29 January, Obokata and her colleagues at the RIKEN Center for Developmental Biology (RIKEN CDB) in Kobe, Japan, and at other institutions in Japan and the United States described an astonishingly simple alternative. All it took to make pluripotent stem cells, they reported, was briefly bathing blood cells from newborn mice in a mildly acidic solution and then tweaking the culture conditions.
Within days, questions about images in the two papers started appearing on blogs and the PubPeer website. RIKEN CDB's Tokyo-based parent organization, RIKEN, launched an investigation on 13 February. New allegations about these and previous Obokata papers continued to emerge.
On 14 March, the investigative committee chair and other RIKEN officials announced the interim results at a standing-room-only press conference in Tokyo. RIKEN president and Nobel laureate Ryoji Noyori opened the event with a deep bow. “I apologize for the great trouble and concerns caused to so many in society by the STAP papers published in Nature by RIKEN researchers,” he said.
The investigating committee concluded that there had been “inappropriate handling of data” for two of the items under investigation, but these were “not judged to constitute research misconduct,” according to a RIKEN press release. But the officials emphasized that they are relying on the scientific community to ultimately decide whether STAP cells can be derived following the method Obokata and her colleagues described. "The mission of the investigating committee is to determine whether or not there has been misconduct; whether STAP cells exist is something for the scientific community to determine," said committee chair Shunsuke Ishii, a RIKEN molecular geneticist, at the press conference, which lasted a marathon 4 hours. He added that to the best of their knowledge, no outside group has reported success in generating STAP cells.
Among the authors, only Charles Vacanti, an anesthesiologist and tissue engineering specialist based at Brigham and Women's Hospital in Boston, says his lab can make STAP cells. Obokata worked under Vacanti for several years, starting in 2008, and developed the STAP technique based on some of his earlier work. Another co-author, Teruhiko Wakayama, who moved in 2012 from RIKEN CDB to the University of Yamanashi, Kofu, confirmed in an e-mail to Science that he had made STAP cells while working alongside Obokata at RIKEN but has not been able to reproduce them at his new lab.
Two other co-authors, Hitoshi Niwa and Yoshiki Sasai of RIKEN CDB, had not attempted to generate the cells independently in their labs before the papers were published. “I’m confused,” says Hans Schöler, a stem cell researcher at the Max Planck Institute for Molecular Biomedicine in Münster, Germany. Niwa and Sasai are highly regarded scientists, he says, and their names as authors lent the papers substantial credibility. RIKEN CDB Director Masatoshi Takeichi said at the press conference that Niwa’s lab is now working to reproduce the results independently of Obokata.
Wakayama’s own doubts have grown. A stem cell biologist known for work cloning mice, he says he was brought onto the team to produce the chimeric mice described in the paper. He used cells provided by Obokata, he says. "I want to believe" in STAP cells, Wakayama says. However, he says, the photos that convinced him that the cells were real are those now believed to have come from a completely different experiment reported in Obokata's doctoral dissertation. "I do not think this is a simple mistake," Wakayama says, adding, "I no longer know to what extent Obokata’s story is the truth." On 10 March, Wakayama called for the papers to be retracted, at least temporarily, until errors are corrected and results confirmed.
At the press conference, Ishii said that the investigating committee had resolved two of six specific issues in the papers. One problematic image is an artifact of image compression and “not falsification or improper conduct,” the interim report states. But the claims that figures in the article were used in Obokata's thesis seem to be on target. Obokata and one of her co-authors told the committee that these figures were used by mistake. Ishii emphasized that this issue is still under investigation and any judgments about misconduct will come after further deliberation.
In a written statement released to the press, Obokata, Niwa, and Sasai apologized for the confusion resulting from the uncertainties and inaccuracies in the papers and wrote: “We are contacting other co-authors regarding the possibility of retracting these papers.”
Vacanti is reluctant. "In the absence of compelling evidence that the data presented is incorrect, I do not believe that the manuscripts should be retracted," he wrote in a statement.
A Nature spokesperson told Science by e-mail that the journal's own investigation “is still in progress.” Not all the authors would necessarily have to agree to a retraction, the spokesperson wrote.
The Nature papers are not the only Obokata publications under fire. Allegations surfaced last week that large portions of one chapter of her doctoral thesis, submitted to Waseda University in 2011, appear to have been copied from a National Institutes of Health website and that footnotes to other chapters were copied and pasted from other publications. Local media have reported that Obokata said she intends to withdraw the dissertation. A Waseda public relations official confirmed that a faculty member had received such an e-mail but said that the university, which is investigating, has not received a formal request from Obokata. If the thesis is retracted, Obokata will lose her doctoral degree.
The glare of the spotlight is not always comfortable.
*Correction, 19 March, 10:12 a.m.: This item has been corrected. An earlier version incorrectly stated that six labs were involved in the papers, but several more contributed to the published work.