原创 Sloppy method tarnishes stem cell research in Japan

Junko Yoshida analyses the scandal surrounding Japanese scientist Haruko Obokata of the Riken Centre for Developmental Biology. Yoshida stresses that implicit trust should have prevailed between Obokata and Riken to discourage her from cutting corners that jeopardised her research.

Front-page headlines, tweets, TV news shows, and tabloids in Japan this week were agog about a Japanese researcher accused of fabricating parts of scientific papers hailed as breakthroughs in stem cell research.

You've probably heard about it. I heard because I happened to be in Japan this week. The Japanese news media's wolfpack instinct of attacking a stray female caught out in the open was remarkable. Everything I despise in the mob mentality of the Japanese press and its audience was on full display.

Still, the story captivated me.

The scientist persecuted in this still-unfolding story is Haruko Obokata of the Riken Center for Developmental Biology in Kobe. She led a team who reported that a simple acid bath might turn cells in the body into stem cells -- except, not so much.

Obviously, it's hard for anyone to resist a story about the downfall of a precocious researcher hailed as a national hero only several weeks before. But as it unfolded, the story exposed a number of important issues fundamental to science and the politics of science. Most importantly, it posed questions about the apparent lack of research rigor in Japan.

The journal Nature published the research results in January. Besides the Riken Center, participants in the study included Brigham and Women's Hospital in Boston and Harvard Medical School. Shortly after the alleged breakthrough, a flurry of comments questioning whether Obokata was able to replicate the stimulus-triggered acquisition of pluripotency (STAP) stem cells surfaced in the global scientific community (and the social networks here in Japan).

The suspicion was triggered by Obokata's sloppy methodologies -- some careless, others bordering on unethical. An image showing a genome analysis appeared to have been spliced together. Images of two placentas from two different experiments looked almost identical. Two more images appeared to have been duplicated from Obokata's 2011 doctoral dissertation, even though the study said they were derived from an entirely different experiment.

Obokata's employer, Riken, a semi-governmental research institute, began investigating the matter in March. Last week, it announced that it had found her guilty of fabrication, and it apologized for behavior that had damaged the credibility of Japan's scientific community.

What did the news conference prove?
On Wednesday, Obokata held a news conference and apologized for her mistakes, which she said were due to her inexperience. However, she denied that she had acted deliberately.

"More than anything, the experiments properly took place," she told reporters. "As the data actually exists, I want to clarify that I did not create these papers with ill intent." She also said she was able to replicate STAP stem cells more than 200 times.

The live TV broadcast of Obokata's nearly two-hour-long press conference, packed with reporters and photographers, was… entertaining at best, not for new discoveries, but as theatre.

As I watched, I was disappointed that Obokata didn't come prepared to offer any fresh evidence (raw data) that could have proved that STAP cells exist. The question-and-answer session turned into a mere war of words.

Who has your back?
I can't help but wonder why none of Obokata's co-authors showed up to face the press. Reportedly, one Japanese co-author is even proposing to retract the study.

Another co-author, Charles Vacanti of Harvard Medical School, told The New York Times, "I continue to feel that the findings presented in these papers are too significant to disregard based on relatively minor errors or external pressures." He also said he'd talk to the other authors before making a recommendation.

That makes sense to me. Why are the rest of the co-authors keeping mum? Why is nobody standing by his own work?

As much as I lament Obokata's sloppiness, which tarnished her research results, I also question the hasty investigation carried out by her elders at Riken and their apparent eagerness to wash their hands and move on. The Japanese institution announced last week that it plans to conduct its own research on STAP stem cells (most likely without Obokata). However, for the immediate future, it seems to have no interest in tracing whatever evidence is still stored in the lab to prove or disprove that Obokata's STAP cells actually exist.

This doesn't seem logical.

To all appearances, Riken is practicing, not science, but the politics of science (not to be confused with political science). As I noted earlier, this is a semi-governmental research institute. Known as a researchers' paradise in Japan, Riken needs to secure budgets from the government. Perhaps this explains its skittishness. But that's no excuse for compounding Obokata's lack of rigor with a less than rigorous investigation.

In the end, in any scientific and engineering research, it's clear that peer review -- buttressed by an established code of conduct -- is the gold standard. The convenience of cherry picking favourable data shouldn't trump hard evidence.

At the same time, it's my belief that, whether you are a researcher or an engineer, you need to know that your employer will have your back. Otherwise, it's hard to stick your neck out, even when that's the right -- and possibly historic -- thing to do.

Riken failed in that test. Perhaps, if the culture there had been healthier, the implicit trust that should have prevailed between Obokata and her employer would have kept her from cutting the corners that undercut her research.


By Junko Yoshida
EE Times